CA1341348C - Human tumor necrosis factor - Google Patents

Abstract

A human physiologically active polypeptide, human Tumor Necrosis Factor (human TNF), comprising a specific amino acid sequence of 155 amino acid residues. The base sequence of the DNA coding for the human TNF has been determined using rabbit TNF cDNA. The human TNF
can be advantageously produced on a large scale by recombinant DNA technique. The human TNF of the present invention has been found to be excellent in inducing necrosis of tumours with no toxic effect upon the a normal tissues of the living body.

Description

w This invention relates to a deoxyribonucleic acid (herein-after referred to as "DNA") coding for a novel human physiolo-gically active polypeptide. This invention also relates to a replicable recombinant DNA containing the DNA, a microorganism or cell transformed with the replicable recombinant DNA, a novel human physiologically active polypeptide obtained by expressing the DNA, a substantially pure human physiologically active poly-peptide having the amino acid sequence described herein, pharma-ceutical compositions containing the physiologically active polypeptide as the effective ingredient, and a process for producing the human physiologically active polypeptide. More particularly, the present invention is concerned with a DNA
coding for human TNF {Tumor Necrosis Factor), human TNF having an amino acid sequence deduced from the base sequence of the DNA, a grocess for producing human TNF from the DNA utilizing recombinant DNA technology, and the use of the product obtained by the process.
In the present specification, amino acids and peptides are represented using abbreviations, as indicated below, approved by IUPAC-IUB Commission on Biochemical Nomenclature {CBN). Inci-dentally, with respect to amino acids and the like having isomers, those ,represented by the following abbreviations are of the L-configuration unless otherwise specified.
Gln: glutamine residue Asp: aspartic acid residue Pro: proline residue Tyr: tyrosine residue Val: valine residue Lys: lysine residue Glu: glutamic acid residue Ala: alanine residue Asn: asparagine residue Leu: leucine residue Phe: phenylalanine residue Gly: glycine residue His: hist:idine residue Ser: serine residue Thr: threonine residue Ile: isoleucine residue Trp: tryptophan residue Arg: arginine residue Met: methionine residue Cys: cysteine residue Polydeoxyribonucleotides and oligodeoxyribonucleotides are represented by sequences of deoxynucleotide residues which are abbreviated as follows:
A: 2'-deoxyadenylic acid residue C: 2'-deoxycytidylic acid residue G: 2'-deoxyguanylic acid residue . T: thymidylic acid residue Unless otherwise specified, the left end of the sequence of deoxynucleotides is the 5' end.
There are known various substances having a capacity for stimulating the reticuloe ndothelial system, for example, physiologically active substances having antitumor activity which are induced by various Gram-positive bacteria and endotoxins. Specifically, Carswell et al discovered that the serum from CD-1 Swiss mice infected with bacillus Calmette-S Guerin (BCG), and after two weeks, followed by intravenous injection of endotoxin has cytotoxic activity against cultured L cells and also discovered a phenomenon that it induces hemorrhagic necrosis of transplanted Meth A sarcoma in the (BALB/c x C57BL/6)Fl mouse. They gave the name of TNF (Tumor I0 Necrosis Factor) to the active substance in the serum [Proc.
Nat. Acad. Sci. USA, Vol. 72 (No. 9), pp. 3666-3670 (1975)].
Thereafter, Ruff et al reported that the rabbit TNF prepared :, according to the above-mentioned method proposed by Carswell et al was purified about 2.000-fold over the serum (J. Immunol., 15 Vol. 125 (No. 4), pp. 1671-1677 (1980)]. Further, Matthews et al reported that the rabbit TNF was purified about 1,000-fold aver the serum [Br. J. Cancer, Vol. 42, pp. 416-422 (1980)].
However,.in Ruff et al and N:atthews et al, the tumor necrosis effect with respect to the purified TNF is not confirmed 20 in animal experiments.
Japanese Patent Application Laid-Open Specification No.
57-140725 (1982) discloses a process for isolating and purifying a proteinaceous physiologically active substance having anti-tumor activity, which is induced by administering to a mammal 25 such as mouse, rabbit or guinea pig at least one substance having a capacity for stimulating the reticuloendothelial system and then injecting endotoin from a Gram-negative bacterium into the mammal, or by adding endoto;:in from a Gram-negative bacterium to a tissue culture containing activated macrophages from a mammal. In this Japanese S Patent Application Laid-Open Saeci~ication, there are also disclosed the molecul4r weight and isoelectric point of the purified proteinaceous physiologically active substance (molecular weight, 39,000 ~- 5,000 as measured by gel filtra-tion and SDS-polyacryl2mide gel electrophoresis; isoelectric point, pH 3.9 ~ 0.3 as measured by isoelectric focusing) but not any detailed structure of the proteinaceous physiologically active substance.
Meanwhile, riatthews reported that there is obtained a substance having cytotoxic activity against L cells by a proc'ss in whica BCG is injected into a rabbit and mononuclear phagocytes prom various tissues of the rabbit are obtained two weeks after the injection, followed by addition of encotoxin _to the cell culture or' the mor_onuclear phagocytes (Br. J. Cancer, Vol. 44 (3), pp. 418-424 (1931)]. Hocaever, in his report, the detailed structure of the obtained substance is not disclosed and, further, there is no evidence showing that the obtained substance is identical with TNF
found in the serum.
Further, there are a number of printed publications reporting that factors having TNF-like bioactivity or a bioactivity similar to that of TNF. For example, Reed et al found such a factor in racrop'nages and the like present in human s peripheral blood [J. Immunology, Vol. 115, p. 395 (1975)], Matthews et al in'leukemia cells derived from human peripheral blood monocytes or from a patient suffering from myelogenous monocytic leukemia [Immunology, Vol. 44, p. 135 (1981)], D.
Barbara in human B cells transformed with Epstein-barr virus [Proc. Nat. Acad. Sci. USA, Vol. 80, p. 5397 (1983)], and B.
Bharat in lymphoblastoid 1788 cell line. The above-~t~ntioned factors are also disclosed in Japanese Patent Application Laid-Open Specifications Nos. 58-15921 (1983), 58-21621 (1983), 1U 58-107197 (1983) and 58-225024 (1983), and British Patent Application Laid-Open Specifications Nos. 2,106,117 and 2,117,385. However, the cells or cell lines capable of efficiently producing such factors have not yet been found.
FurthAr, kith reS_~~.eCt t0 Such factOrS, there are many m3tterS t0 be elucidated such as their structures and properties'.
Ito [Japanese Patent Application No. 58-251817 (1983)]
made studies on properties and structure of rabbit TNF and rabbit-TNF producing cells. As a result, he obtained cells capable of producing a substance having cytotoxic activity against I. cells by administering a substance having a capacity for stimulating the reticuloendothelial system to a rabbit, followed by injection of endotoxin derived from a bacterium -r~~~c~
~'~_= into the rabbit, and then obtained such a substance usingW
cells. He also affirmed that the molecular weight and immunological properties of the substance having cytotoxic activity against L cells obtained using the above obtained 134134a cells are in agreement with those of TNF obtained from rabbit serum. Meanwhile, with the progress of genetic manipulation techniques, it became possible to determine the structure of a protein so long as a DNA coding for the protein is obtained in isolated form. This is so because the structure of the isolated DNA can be determined and, then, the structure of the protein can be deduced from the structure of the DNA. Further, it became possible to produce a protein from a DNA coding f or the protein utilizing ~a microorganism or cell culture. Ito applied the above-mentioned genetic manipulation techniques to the cells capable of producing a substance having cytotoxic activity against L cells. As a result, he succeeded in isolating a DNA coding for rabbit TNF, determining the structures of the DNA and rabbit TNF, and producing rabbit TNF using the D2dA.
As is apparent from the foregoing, Ito has made a great success in producing rabbit TNF. However, it should be noted that the administration of TNF to an animal which is not 2Q the origin of the TNF has a danger that anaphylactic shock might be caused. This is so because TNF is a polypeptide and, hence, when TNF is administered to an animal which is not the origin of the TNF, the TNF functions as an antigen to produce an antibody. For this reason, when TNF is intended to be administered to a human body, the use of TNF derived from human beings is highly pref enable. However, even the structure of human TNF has not been successfully elucidated. Therefore, _ 7 _ the det~r.,.ination of the s~ruc~ur~ o~ the DNA toting for human TN: has been s~~onalv needed.
T:~e pr'se:~t inventors have made atensive and intensive studies on the st=uc~ura of the DNr toting for human TNF.
As a result, the present inve:~tors have su;prisinaly found that a human polvpeptide gene and a rabbit TNF gene can be cloned by the use of rabbit cDNA as a probe, that the sue' DNA coding for human pol~rpeptide can be skillfully isolated and the structure the=eof can be determined by comparison between rabbit TNF gene, human r~olypeotide Qene and rabbit cDNA with respect to the homology of their base seauences, t'1?~ t the S tr uC tllre O f pLlr a DLI~Tt~ COdInQ f Or hLman pOlypep tide Can be Sivl~lfullV dote=mined and such a Dure DNA C3n be obtained, and that a huma:~ polvpeptide produced using the D~1A coding for the hl::i;an pO1 VDeO tide has a c y to toxic activi ty acainst L cells.
The arose~~t invention has been mace based on sucz novel findi_ncs .
'j.'a'?e. °~Ors , 1 t is an Obi eC t OL the p'="°se:lv lnVelltl0n t0 provide a human phvsiolocically active polypeptide.
T_t i s aro~ze= cbje~~ of the preszn~ invention to provide a DNh coa._nc for human T~iF.
I t i s s ti= i another objet t of the present invention to provide a replica:.le recombinant DNA? comprising a DNA coding for human T~1F and a replicable expression vehicle.
It is a further object of the pr'sent invention to provide a microorganism or a cell transformed with a recombinant DNn of the kind as me.~.tic:~Pd above.
_ g _ It is a further object of the present invention to provide a process for producing a human physiologically active polypep-tide of the kind as mentioned above.
The foregoing and other objects, features and advantages of the present invention will be apparent from the-following Fig. 1 illustrates ~the restriction maps of .~~.$. each containing a DNA coding for a conventional rabbit physiologi-cally active polypeptide;
Fig. 2 illustrates the flow-sheet of the method for the detailed description taken in connection with the accompanying drawings in which:
~ ~ asm~~ d preparation of a recombinant DNA (pTPlF-lac-1) coding for the conventional rabbit physiologically active polypeptide;
Fig. 3 illustrates the flow-sheet of the method for the preparation of another recombinant DNA (pTNF-lacUVS-1) coding for the conventional rabbit physiologically active polypeptide;
Fig. 4 illustrates the restriction map of the portion of a plasmid containing a gene for human physiologically active poly-peptide of the present invention;
Fig. 5 illustrates the restriction map of the portion of a plasmid containing a gene for a conventional rabbit physiologi-cally active polypeptide;
Fig. 6 illustrates the flow-sheet of the method for the preparation of a recombinant DNA (pHTNF-lacW5-1) coding for a human physiologically active polypeptide of the present invention; and Fig. 7 illustrates the flow-sheet of the method for the preparation of another recombinant DNA (pHTi~F-lacW5-2) coding for the human physiologically active polypeptide of the present invention.
_ g -~ 34~ 348 Essentially, according to the present invention, there is provided a human physioloaicallv active pol;rpeptide having an amino acid secuence represented b~~ the following formula (I)~:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn G1n Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly A1a Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn A D a A1 a G1 o r:ly Gl ~ t~yl Tyr ph°
Arcr Pro Asp Tvr Leu so _ h_ .._ a S_.
Gly Ile Ile A1~ Leu wherein Gin stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue.

The human physiologically active polypeptide of the present invention also includes a polypeptide having an amino acid methionine attached to the N-terminus of the above-mentioned amino acid sequence and an intermediate having a partial or entire signal peptide for human TNF attached to the N-terminus of the above-mentioned amino acid sequence. It is possible to change part of the structure of a DNA coding f or a polypeptide by natural or artificial mutation without significant change of the activity of the polypeptide. The human physiologically active polypeptide of the present invention includes a polypep-tide having a structure corresponding to homologous variants) of the polypeptide having the above-mentioned amino acid sequence. All such physiologically active polypeptides are hereinaf ter referred to as "hu:.;an TNF".
In another aspect of the present invention, there is provided a deoxyribonucleic acid comprising a base sequence coding for a human physiologically active polypeptide, said human physiologically active polypeptide having an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly. Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln .ra Giu Thr Pro Glu Gly Ala Glu Ala Lvs Pro Trn T~~r Gl a Pr., Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lvs Gly Asp r':rg Le a Ser A1a Glu Ile Asn Arg Pro Asp Tyr Leu F.sp Phe A1~ Glu Ser Gly Gln Val Tyr Phe Glv Ile Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an ascartic acid residue, Pro a proline residue, Tyr a tyrosine resicue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, 1~ Ile an isoleucine residue, Tro a tryptophan reS'!due, :.T"g an crglnine re~l.~~.t:e, 1":et a methionine residue, and Cys a cysteine residue.
In =urther aspect of the present invention, there is provided a deoxyribonucleic acid comprising at least one base seauence selected from the group consisting of a base seauence represented by the following formula tII) and a ccmplementary base seauence to said base sequence:

TCA TCT TCT CGA ACC CCG AGT GAC AnG CCT GTA GCC CAT GTT GTA
GCA AAC CCT CAA GCT GAG GGG CAG CTC C~:G TGG CTG 3=.C CGC CGG
GCC AAT GCC CTC CTG GCC A.nT GGC GTG GAG CTG AGA G=.T AAC CAG
CTG GTG GTG CCA TC=~ GAG GGC CTG TAC CTC ATC TAC TCC CAG GTC
CTC TTC AAG GGC CA_~1 G::C TGC CCC TCC ACC CAT GTG CTC CTC ACC
CAC ACC ATC AGC CGC .TC GCC GTC TCC TAC CAG ACC AAG GTC AAC
CTC CTC TCT GCC ATC ~aG AGC CCC TGC CAG AGG GAG ACC CCA GAG
GGG GCT GAG GCC A_~G CCC TGG TAT GAG CCC ATC TAT,CTG.GGA GGG
GTC TTC CAG CTG GAG P_~G GGT GAC CGA CTC AGC GCT GAG ATC AAT
CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
GGG ATC ATT GCC CTG
wherein A stands for a deoxvadenylic acid residue, G a deoxvguanylic acid residue, C
a deo~>vc~~=idvlic acid residue and T
thvmicylic acid residue and wherein the left end and riche end of the =c=:.~.ula (II) rep=2sent ~ ' -h_acr oxy l gr pup side and 3' -hydr oxgi cr pup side, respec~ively.
The DNA of the present invention includes a DNA
comprising a base secuence having ATG (A, T and G are as mentioned above) attaczed to the 5'-end of the above-mentioned base seauence in order to produce mature human TNF by means of cultera of a miceoorganism or cell.
The DNA of the present invention also includes a DNA having a ~'-flanking DNA cooing for a partial or entire signal peptide of human TNF.

The structure of a DNA and the structure of the pol.y-peptide deduced therefrom may be partially changed by natural or artificial mutation without causing the main activity of the polypeptide to be changed. Hence, the DNA
of the present invention may alternatively have a base sequence that codes for a polypeptide with a structure corresponding to that of a homologous variant of any of the aforementioned polypeptides.
In accordance with degeneracy of genetic code, it is possible to substitute at least one base of the base sequence of a gene by another kind of base without causing the amino acid sequence of the polypeptide produced from the gene to be changed. Hence, the DNA of the present invention may also have any base sequence that has been changed by substi-tution in accordance-with degeneracy of genetic code. In this instance, the a:-ino acid seauence deduced from the base sequence obtained by the above-:mentioned substitution is identical with the amino acid seauence of the formula (I) as defined before.
In a further aspect of the present invention, there is provided a replicable recombinant DNA which comprises the above-mentioned deoxyribonucleic acid according to the present invention and a repllCable expression vehicle. The recombi-nant DNA is capable, in a transformed microorganism or cell culture, of expressing a polypeptide comprising the amino acid sequence of human TNF. As the suitable vehicle, there may be mentioned, for example, pHTNF-lacUVS-1 and pHTNF~-lacUVS-2 expression vehicles.
Further, the present invention is directed to a microorganism or cell culture transformer with a recombinant DNA capable of e:~pressing a polypeptide comcrising the amino acid sequence of human TNF. Examples of such microorganism or cell culture include Escherichia coli, Bacillus subtilis, yeasts and higher animal cells.
In an even further aspect of the present invention, there is provided a method for producing the human physiologically active polypeptide of the present invention which comprises:
(a) ligating the deoxyribonucleic acid of the formula (II) as defined above to a replicable expression vehicle to obtain a replicable recombinant DNA comprising said deoxyribonucleic acid and said replicable expression vehicle;
(b) transfor.;.ing ells of a r.:icroorgarism or cell culture with said replicable recombinant DNA to form transformants;
(c) selecting said transformants from parent cells of the microorganism or cell culture;
(d) incubating said transformants, causing said transformants to express said deoxyribonucleic acid and produce a human physiologically active polypeptide; and (e) isolating said human physiologically active polypeptide from the incubated transformants.
According to the method of the present invention, the above-described polydeoxyribonucleic acid of the present invention is ligated to a replicable expression vehicle as a vector to obtain a replicable recombinant DNA containing the above-mentioned polydeoxyribonucleic acid. A micro-organism or cell culture is transformed with the thus obtained replicable recombinant DNA to obtain a transformed micro-organism or cell culture containing the recombinant DNA.
The thus obtained transformant is isolated from the parent microorganism or cell culture by means of a phenotypical trait imparted with the DNA. The thus obtained transformed microorganism or cell culture is grown to effect expression of the genetic information that is encoded on the above-mentioned deoxyribonucleic acid, thereby producing a physiologically active polypeptide according to the present invention.
Furthermore, the present invention is directed to a human T1~F, in mature form, secreted from host cells as a direct expression product. As a process for obtaining such a mature human TNF, there may be mentioned, for example, a process comprising constructing a DNA sequence so as to bond an amino acid sequence, known as a signal peptide, composed of from 15 to 40 amino acids that is derived from a microorganism or higher animal to the terminus of the amino acid sequence of the mature TNF.
The human TNF may be obtained as follows:
1. A bacteriophage a/rabbit genomic library and a bacteriophage /human genomic library prepared by Prof. T. Maniatis, Department of Biochemistry and Molecular Biology, Harvard University, 7 Divinity ca rc_, Avenue, Cambridge, Massachusetts 0213E, L3.S.. used.
These materials may be prepared according to the following procedures (see Cell, 15, p.687 (1978)]:
(1) rabbit or human tissues, for eaample rabbit or human pancreas tissue, are reduces to frozen powder and treated to digest RNA and protein materials and provide, on precipitation, high molecular weight rabbit or human DNA;
(2) the high molecular weight DNA is partially digested for random cutting with respect to gene locus;
(3) the resultant DNA fragments ere size-fractionated giving from 15 to 20 kilo base pair (kb) fra~~-its;
(4) the resultant fragments of Step 3 are cloned using a ~~ Charon 4A ohaae vector; and ( S ) the resul tart vec for s are pacr;aged in v i tro to infectious phage particles containing rDNA to obtain the above-mentioned rabbit or human cenornic library.
2. The rabbit TNF cDNA obtained in Reference Lxa:;~ple 3 is 32P-labelle~ by P.W.J. Rigby et al's nick translation method (see J. Mol. Biol. 113, p.237 (1977)1.
3. Each of the bacteriophage /rabbit genomic library and bacteriophage /human genomic library is plated to virtual confluence on a lawn of bacteria and screened for hybridization with the 32p-labelled rabbit TNF cDNA.

4. From the appropriate clones, the corresaonding DNA is isolated, res~riction mapped and analyzed by Southern hybridization [see ...:~I. Southern, J. hiol. Biol. , 98, p.503 (19751].
Restriction fragme.~.ts containing rabbit or human TNF
genes are subcloned into plas:nid vectors and then sequenced.
5. The base sequence of the rabbit TNF cDNA is compared with that of the rabbit TNF gene to determine the exons (certain sequences of bases which code for the amino acid sequence of rabbit TNF) and introns (certain sequences of bases which do not code for the amino acid sequence of rabbit TNF) of the rabbit TNF gene.
6. 11'lere~a.L°r, the base secuence of the human TNF gene is compared with gnat of the rabbit TNF gene to determine tine e:cons and in tr ons of the human TivF gene .

7. The amino acid se~uezce of rabbit TNF that has been deduced ::ron the base sequence obtained by deleting the introns of the rabbit TNF gene and combining the exons thereof is af=firmed to be in agreement with that deduced from the base sequence of the rabbit TNF cDNA.

8. Next, the amino acid sequence of human TNF is'deduced from the base sequence of the DNA coding for human TNF
obtained by deleting the introns of the human TNF gene and combining the e:~ons thereof. The amino acid sequence of the human TNF is affirmed to be partially in agreement crith that of the rabbit TNF.

9. Then, the DNA, coding for human T':F is tailored in vitro for insertion into an appropriate e::pression vehicle to form recombinant DNA containing the coding DNA. The recombinant DNA is used to transform an appropriate host cell which is, in turn, permitted to grow in a culture and to exaress the desired human TNF.

10. The human TNF thus produced has 155 amino acid residues in its mature form, beginning with serine, Vherl it has a signal peptide in its presecuence, the signal peptide is very hydrophobic in character.
The foregoing discloses the procedures for obtaining the human TNF gene, the base seauence of the D~IA coding for human TNF and the process for producing the human TNF by the use of the DNA. However, it should be understood that the foregoing disclosure is not intended to limit the invention and that obvious chances may be made by those skilled in the art without changing the essential characteristics and the basic concept of the invention.
Due to the variable use frequency of a colon (genetic ~0 code? corresponding to each amino acid and for other reasons, a partial or entire portion of the base sea_uence of the DNA coding for human TNF may be substituted by an organic chemically synthesized artificial DNA without causing the amino acid sequence of the polypeptide obtained therefrom to be changed.
Presumably, the human TNF may be intracellularly produced in immature form as a prepeptide or prepropeptide, -.f~ _ ~ b ~ e5x=c~:.
which may be ~cm~ via an-intermediate form to a mature TNF
in the processing stage. The immature form of hu."an TNF
may be de~?uceu from the bass seguence of the human TNF
gene. The TNF DNA comprising a DNA encoding the TNF in immature or intermediate form may also be recombined with a natural or artificially s=~nthesized DNA.
One application of this techniaue may be attained by inserting the methionine colon (ATG) in the 5'-end and inserting at least one stop colon selected from,TAA, TAG and TGA in the c;;~ r 3'-end of the mature or intermediate t~-s immature TNF DNA. Due to the presence of the methionine colon, the mature or inter-mediate or immature TNF may be produced on the mRNA synthesized with the aid of an appropriate promoter. However, the methionine residue attached to the N-terminus of the TNF is cleaved or not cleaved according to the kind of the host cell employed.
The purpose of inserting the stow cocon is to stop translation of the mRNA transcripted from the Ti~F DNA at an appropriate position (C-terminus of poly~eptide of-the formula I).
Another application of this technique may be attained by adding to the DNA a highly hydrophobic base sequence known as a "signal seauence". By this addition, it may become feasible to secrete the TNF to outside the host cell or, in the case of a gram-negative bacteria, into the space known as "periplasm".
When a vector in which a start colon is incorporated is employed, a fused peptide may be produced which consists of the human TNF and a peptide attributed ~o the vector.
In this case, the fused peptide may be cleaved chemically or enzymatically. Alternatively, the fused peptide, if ~ 34~ 348 the main activity of the human TNF is not adversely affected, may be used as it is.
The human T~'F DNa may be connected, at its region upstream of the S'-end, to the gene sequence of a promoter thereby to obtain a T:zF DNA-promoter sequence which does not hinder its replic,_--..tion anc does not cause translation of the resultant P,NA to be adversely affected. The thus obtained TNF DNA-promoter sequence may be comr:ined with a vector which is replicable in a bacterium or higher organism cell to obtain a recombinant gene. The thus obtained recombinant gene may be used to transform a bacterium or higher organism cell used as a host. The thus obtained trap sformant may be cultured to effect e:cpression of the TNF
gene in order to produce the human T:~F.
Vher. Escheric:iia coli is used as the above-mentioned host, there may be mentioned, as the suitable host, various mutant strains of E. coli K-12, such as HB101(ATCC 3369x).
C600K(ATCC33955) . D1210, RRI (r.TCC313x3) , f!C1061, T_.E392 ':Sl'~~13 (ATCC33572), JM101 (ATCC33876)~and X1776 (ATCC3124x).
When the E. coli host is emaloyed, there may be mentioned, as the suitable vector, plasmids such as pBR322, pBR325, pBR327, pUC3, pUC9, pMB9(~TCC37019), pJB8(ATCC37074) and pF:C7 (ATCC3708x ) , a. phages such as ~ gt, ~~ B and Char on xA, and M13 phage. To have TNF produced in the E. coli cell, a promoter selected from the promoters of the E. coli and phage genes may be employed. Examples of the suitable promoter include the genes for lactose degradation enzyme (LAC), UV5 mutant thereof, penicillinase (BLA) and tryptophan synthetase (TRP), ~ phage PL promoter and tac promoter which is a fused promoter of try_~tophan synt:~etase and lactose degradation enzyme.
When Bacillus subtlis is used as the host, there may be mentioned, as the suitable host. BD170 strain (ATCC33608), BR151 strain (ATCC33677) and MI112 strain (ATCC33712). When the Bacillus subtilis host is employed, there may be mentioned, as the suitable vector, plasmids pC194(ATCC37034), pUB110(ATCC37015), pSA2100(ATCC37014) and pE194. Further, when the Bacillus subtilis host is employed, there may be mentioned, as the suitable promoter, the genes Tor Chl C~'~:.~.p~le~=COl aC°tylati Gn enz:t:«e (~ AT) , pe-liC~l 1 iilaS~ and anti-erytzromycin.
When a yeast is used as the host, there may be mentioned, as the suitable host, stains of Saccharomvces c~revisiae such as RH218 (ATCCa4076) , SHY1 (:.TCC44769) , SH~.'3 (ATCC44771) , D131A, X83 and 830. when the yeast host is employed, there may be mentioned, as the suitable vector, plasmids such as YEpl3(H.TCC37115), YEp6, YRp7 and YIpS. Further, when the yeast host is employed, there may be mentioned, as the suitable promoter, the genes zor acid phosphatase, alcohol de:~ydrogenase (r.DHI) , tryptop han synthetase (TRP) , phosphoglycerate kinase (PGn). cytochrome B(COB) and actin.
When a higher organism cell culture is used as the host, there may be mentioned, as the suitable host, the cell cultures o~ monkey kidney, COS and mouse C127(ATCC 1616).
When the higher organis:~ cell culture host is e:r;ployed, ./' i.t ~
there may be mentioned, as the suitable vector, SV40~~a~-~ae~-ire-°p~a:~:~~x~.

~ 34~ Sae The novel human physiologically active polypeptide of the present invention incuces necrosis of tumors with no toxic effect upon the normal tissues of the living body.
The active polypeptide of the present invention may be formulated according to known methods to prepare pharma-ceutical compositions which are useful for the inhibition of cell proliferation, e.g. malignant tumor~cells prolifer-ation. The active polypeptide may be combined in admixture with a pharmaceutically acceptable carrier vehicle.
An effective amount of the active polypeotide of the present invention may be mixed with a suitable amount of vehicle in order ~t0 prepare phar~rtaceuticaliy acceptable compositions suitable for effective administration to the reciaient.
The physiologically active polypeptide of the present invention may be administered, to subjects recuiring anti-tumor or antiviral treatment, as an injection, eye drop;
nasal drop, inhalant, external preparation, oral acministra-tion, rectal administration or vaginal tablet. The daily dose of the polypeptide of the present invention per adult may be generally in the range of from 50 to 100,000,000 units. It may be preferably in the range of from 50 to 500,000 units in the case of local administration, from 1,000 to 1,000,000 units in the case of general injection such as intravenous injection and intramuscular injection, and from 10,000 to 100,000,000 units in the case of oral administration. The daily dose may be increased or deer eased accorC.ina to the direr Lion for use and s~:; ;Ntom of recipient.
The terminology "1 unit" used above means a quantity of the physiologically active polypeptide of the present invention by which 50 0 oL 1 x 105 cells/ml of L-~? cells (Pmav~-icar: Type Culture Collection CCL 1.2) are killed. The above-mentioned quantity is measured as follows. As culture vessels, there are employed 96-well microtiter plates produced by Flow Laboratories, Inc. (U.S.A.), and L--:~. cells are cultured in Eagle's minimum essential medium containing 1 v/v o of bovine fetal serum [the composition or this medium is described, for example, in Tissue Culture, edited by JuiaiaoSilke ~ariai et al, Asakura Shoten, Japan (1967) ] . A
sar,~ple (0.1 ml) serially diluted with the medium and the L-M
cell suscension (0.1 m1, 1 x 105 cells/ml) are mixed into each well of the plates and the plates are incubated at 37°C for 48 hours in air containing 5 o carbon dioxide.
At the end of the cultare period, 20 ~l of glutaraldehyde is added to fix the cells. After fixation, the plates are washed with distilled water and allowed to dry, and 0.05 %
rnethylene blue (0.1 ml) is added to stain the viable cells.
The plates are thoroughly washed with distilled water to remove excess dye and allowed to dry. 0.36 N Hydrochloric acid is acded to each well to extract the dye from stained cells. Absorbance of each well at 665 nm is measured with .Titertek Multis::an produced by Flow Laboratories, Inc, The absorbance is proportional to the number of viable cells.

The above-mentioned quantity of the physiologically active polypeptide of the present invention by which 50$ of 1 x 105 cells/ml of L-M are killed is obtained by plotting the dilution versus the absorbance on a graph.
The physiologically active polypeptide of the present invention may be suitably administered parenterally.
In the parenteral preparation, there may be incorporated as an additive, a thickener such as sucrose, glycerine, methylcellulose, carboxymethylcellulose or the like and/or a pH adjusting agent such as various inorganic.
salts. The polypeptide may also be suitably administered in the form of a tablet. In the tablet, there may be incorporated, as an additive, a vehicle such as starch, lactose or the like.
As a result of animal experiment, it has been found that a mouse tumor is completely healed by one or two injections only, in most cases. In particular, an aliquot of artificial neoplastic cells (Meth-A cells) were transplanted to the skin of each mouse. When the tumor grew to have a diameter of 6 to 7 mm, as little as 0.6 ug of the polypeptide of the present invention was injected. A week later, a scab appeared. Two weeks later, hairs began to grow, which means complete healing of the tumor. Later, pregnancy and successful birth were observed for the mice.

The present invention will be described in more detail with reference to the following Referential Examples and working Examples, which should not be con-strued to be limiting the scope of the present invention.
In practicing the present invention, construction of a recombinant DNA and insertion of a recombinant DNA to G~ fe~
"a microorganism ~ carried out in accordance with the procedure described in the following experimental reports (Literatures (1) to (4)], unless otherwise indicated.
(1) Yasutaka Takagi, rlanual For Genetic Engineering, Kodan-sha, Tokyo.
(2) Yasutaka Takagi, Experimental _M.ethod In Genetic Engineering, Kodan-sha, Tokyo, (3) T. Maniatis, E. F. Fritsch, J. Sam Brook, Molecular Cloning, Cold Spring Harbor Laboratory, New York.
(4) Ray Wu et al., Method in Enzymology, Vol. 101, Academic Press, New York.

Abbreviations used in Referential Examales and ~:;ar~:.les rIOP~S: morpholinoproaanesulfonic acid LB medium: Luria-Bertani medium DMSO: dimethylsulfoxide PFU: plaque forming unit EDTA: ethylenediaminetetraacetic acid SDS: sodium dodecyl sulfate BRL: ' Bethesda Research Laboratories Inc.

DMT: dimethoxytrityl lac: lactose Tris: tris(hvdroxvmethvl)aminomethane X_~R-5: - X-ray 'il~: manuLactured and sold by Eastr"aa Kodak Company,. U.S.A.

1 x SSC: 0.15 M NaCl + 0.015 sodium citrate, pH7 2 x SSC: 0.30 ri NaC1 + 0.030 M sodium citrate, pH7 3 x SSC': 0.~5 M NaCl + 0.05 M sodium citrate, pH7 S x SSC : O . % ~ hI N'aCl + O . O7 5 M SCdlum C'_ tra to , pH7 6 xSSC: 0.9 bi NaCl + 0.09 M sodium citrate, pH7 ~ FDSS:

50 s deionized formamide + 5 x Den:~ardt's +

5 x SSPE + 0.1 o SDS + 100 ~g/ml denatured calf thymus DNA

_ 27 -1 341 34 g SSPE: 0.18 M NaCl ~ 10 mM Na:~PO~
1 m~1 EDTa, pH~.4 SM: phage storage medium :ohich contains 5.8 g of NaCl 2 g of MgSp4~7H20, 50 ml of 1 t~i Tris~C1(pH7.5) and 5 ml of 2 ~ gelatin per liter NZ-broth: medium which contains 10 g of NZ amine, 5 g of NaCI and 2 g of MgS04~7H20 (NZ amine is a Type-A hydrolysate of casein manufactured and sold by Humko Sheffield Chemical Division of Kraft, Inc., U.S.A.) IPTG: isopropyl t:lioo~alactoside J~,.,J , ~
~.~brc~rno .. cf _~~l~f'"c~ '3 ~ Il?~~ ~ ~L~aG~p 51~~.
x-gal: -5-~.~~_4_rh,l,or..i .-TAE: 0.04 M Tris-acetate (pH8.0) - 0.002 i~I EDTA
5 x Denhardt's solution: an aqueous solution containing Ficoll 1000 mg, polyvinyl-pyrrolidone 1000 mg and BSA
1000 mg per ~1 i ter bp: base pair 1 341 34 $
Referential Example 1 (Evaluation of cytotoxic activity against L cells) The evaluation of the cytotoxic activity of the physiolo-gic ally active substance prepared in the following Referential ales and ales against L cells is effected by measuring its cytotoxic effect on the L929cells (American Type Culture Collection CCL 1), in accordance with the method of Ruff et al [see Lymphokines, Vol. 2, edited by E. Pick, Academic Press, N.Y., 235 (1980)]
~, i~ 7~T
or the method described in J. Irnmunol, 126, - (1981)].
The method of evaluation of the cytotoxic activity of the physiologically active substance prepared in the Examples is explained below.
As culture vessels, there are employed 96-well microtiter plates produced by Flow Laboratories, Inc. (U.S.A.), and L929 cells are cultured in Eagle's mi imum essential Cj ~' ~t~ ~ c c ~ l ~ S er~ cd r1r'a medium containing 1 v/v ~qand 5 ug/mQ (f.inal concentration) of actinomycin D [the composition of d ___ this medium is described, for example, in Tissue Culture, edited by Junnosuke Nakai et al, Asakura Shoten, Japan (1967)].
A sample (0.1 ml) serially diluted with the medium and the L929 cell suspension (0.1 ml, 1 x 105 cells) are mixed in each well of the plates and the plates are incubated at 37°C
for 21 hours in air containing 5 o carbon dioxide. At the end of the culture period, 20 ul of a 20 o aqueous solution of qlutaraldehyde is added to/ fix the cells. After fixation, the plates are washed with distilled water and allowed to dry, and 0.05 ~ methyler_e blue (0.1 ml) is added to stain the viable ells. The alates are thorouahlv washsd with distilled Ovate= to remove e::cess dve and allowed to dry. 0.36 N H;~crochloric acid is added to each well to extract the dye from stained cells. Absorbance of each well at 665 nm is measured with Tite~t~!: b:ultiskan (produced by Flow Laboratories;- Inc., U.S.A.).The absorbance is proportional to the number of viable cells. The cy totoxic activity of the physiologically active substance, unit/ml, is defined as the reciprocal dilution of the physiologically active substance that causes 50 ~ cytotoxicity, and cari be obtained by plotting the dilution versus the absorbance on a graph. The "1 unit"
used in Referential Examples means a quantity of the rabbit TNF by which 50 ~ of 105 cells/ml of L929 cells are killed.
On the other hand, the amount of protein is determined by a method in which Coomassie Brilliant Blue 6250 is bonded to protein, according to the teaching of Bradford et al [see Anal. Biochem. Vol, 72, pp 2a8-254 (1°76)].

Referential Exa.-,:ale 2 S t.ep 1 (Preparation of TNF from rabbit serum) Female rab~its, weighing 2.5 to 3 kg, are injected with 50 mg of formalin-killed Pronionibacte=ium acnes (Corvne-bacterium parvu:~; Wellcome Research Laboratories, England) through the ear vein. Eight days later, 100 ~g of endoto;cin (lipopolysaccharide from Escherichia coli 026:86, produced by Difco Laboratories, U.S.A.) is injected again through the ear vein and 2 hours later whole blood is collected from the heart. To the collected blood, heparin sodium is added in an amount of 100 units per 100 ml. The blood is then centrifuged while coolinc at 5.000 rpm for 30 minutes to remove blood cells and insoluble solids. As a result, a plasma ( 2 . 4 1 i per s ) havinc a see um TNF cv totor:i c ac ~ivity of 3 x 104 units; m1 is obtaine~? _rom 40 rabbits.
S yep 2 (?artial puriFication of TNF from rabbit serum) To the plasma (2.4 liters) obtained in Step l,~added 24 g of cellite. The resultant is stirred for one hour, and then subjected to filtration. The filtrate is mixed with 1.2 liters of 0.04 M Tris-HC1 buffer (pH 7.8).
and then applied to a column of DEAE-Sepharose CL-6B (manufac-..
Lured and sold bv_ Pharmacia Fine Chemicals, Inc. Sweden) C~. C~ a f..
sufficiently eaui librated with ~-:9'4' M Tris-HC1 buffer (pH
. 7.8) containing 0.1 M NaCl. The column is washed with 0.04 M Tris-HC1 buffer, and the adsorbed TNF is eluted with .~' ,-i' /'~~C' fna. r/~, 0. 04 bI Tris-HC1 buffer (pH 7 . 2 ) containing 0. 18 ri P~aCl.
Fractions exhibiting cytotoxic activities against L cells are concentrated bv_ ultrafiltration. The so obtained 7~. S. _ ~ ~ (.~
~ _° concentrate is applied to a column of Sephacryl '~~-A~6~9-(manufactured and sold by Pharmacia Fine Chemicals, Inc.
Sweden) sufficiently equilibrated with S mM phosphate buffer and gel-filtered using the same buffer. The active fractions are concentrated by ultrafiltration, whereby a purified TNF
having an activity of 3.5 x 106 units and a specific activity of 18 x 106 units/mg is obt,ined.
Step (Anti-TNF antibody) The rabbit serum TNF partiallw purified in Step 2 is mixed with complete Freund's adjuvant (1:1), and then injected subcutaneously at the back of a 12 week age BALB/c male mouse.
The above operation is repeated 2 and 4 weeks after the initial injec~ion. One week after the last injection, whole blood is collected. From the collected blood, a serum is obtained.
The so-obtained serum is. added to the culture medium for evaluation of the cytotoxic activity of TNF against L cells in such an amount that it is diluted 500-fold in final concentration. The cytoto~:ic activity of the rabbit serum TNF against L cells is evaluated in the same manner as described in Referential Exarnole 1. It is found that the rabbit serum TNF exhibits no cytotoxicity against L cells.
From the above result, it can be concluded that the mouse °~ ~1'Yl-~~' ~L~~ .

1 341 34$
serum obtained in this step contains an antibody to the rabbit seism TNF (hereinafter reTerre.:. to as "anti-TNF
antibody").
Referential Example 3 Step 1 (Preparation of TNF-producing cells) A female rabbit is injected intravenously with formalin-killed cells of Prooionibacterium aches (Coryne-bacterium a~ rvum; Wellcome Research Laboratories, England).
Seven days later, the rabbit is subjected to tracheotomy, and the lung ~is washed with a physiological saline solution, whereby floating cells are obtained. The so obtained cells are washed with a physiological saline solution. Using as a culture medium RPf~II 16.0 (Flow laboratories Inc., U.S.A.) containing 10 v/v o fetal calf serum, the cells are incubated at 37°C in air containing 5 o carton dioxide.
The cell culture is divided into two groups, and to one Esc ~ r ~ ~ h '~ GC
of them endotoxin derived from coli (lipopolysaccharide from Escherchia coli 026: E6, produced by Difco Laboratories, U.S.A.) is added at a concentration of 10 ~g/ml. The same amount of stsrile water is added to the otter. The supernatant of the cell culture to which endotoxin is added exhibits cytotoxic activity against L

cells, and the activity reaches the maximum value within seven hours. Such activity is dissiaated by the anti-TNF
antibody, but is not dissipated by the normal mouse serum.
On the other hand, the supernatant of the cell culture to which no endo toxin is' added eazibi is no cy to to:~icity against L cells.
S teo 2 (Molecular weight oy TNF) To the cell culture prepared in Step 1 to which endotoxin is added, radioactive L-[35Sj methionine (1300 Ci/mmol, produced by Amersham Industries plc, England) is further added (1 mCi/ml). In accordance with the method oz Laemmli [see Laemmli, U.K. (1970), Nature (London), Vol. 227, pp 680-685j, the supernatant is analyzed by the SDS-polyacrylamide gel electrophoresis. The gel concentration is adjusted to 12.5 wt ~. After the electrophoresis, the gel is treated with ENHANCE~ (trademark of a product of New England Nuclear Inc., U.S.A.), and after dr~~ing, is exposed to X-ray film (Fuji RX, manufactured and sold by Fu3i Photo Film Co., Ltd., Japan). In the supernatant of the ~ ~-~.~ ~~t~

cell culture in the presence of endotoxin, it is observed that a substance having a molecular weight of about 17500 is formed.
Further, the supernatant of each cell culture prepared in Step 1 is subjected to SL:S -polyacrylamide gel electro-phoresis in the same manner as described above. There-after, the gel is shaken in 2.5 $ hP a0° (a surface active agent sold by Calbiochem, U.S.A.) for one hour, and then in water for two hours. After shaking, each migration lane is separated by cutting, and cut into strips of 2 mm-width in a direction perpendicular to the direction of migration. Each strip is cultured with L cells, and evaluated for cytotoxic activity against L cells. In the lane on which the supernatant of the cell culture containing endotoxin is develoged, cytotoxicity against L cells is observed at a position corresponding to the molecular weight of 17500. No cytotoxicity is observed at other positions.
Step 3 f~X~~rQC~IC>r1 ;~ (C-~;~aof mRl~iA) The cell culture as prepared in Step 1 is incubated f or 2 hours of ter addition of endotoxin, followed by centri-fugation to collect cells. Extraction of cytoplasmic RNA from the collected cells and e:~:traction of mRi~i,? from the cytoplasmic RNA are e~fected in acecr~~ance with the method of Chirgwin et al (see Chirg~~rin, ,J.t~l. et al, Biochemistry, Vol. 18, p. 5294 (1°79)3. 4 ml of a 4 r1 guanidine thiocvanate solution is added to 3 x 108 cells, and the mixture is pulverized by means of a homogenizeY~ (riodel: P~i-7, manufactured and sold by Nihon Seiki Seisakusho, Japan). The residues are removed by centri=ugation, and 2.4 g of cesium chloride is dissolved therein. The mixture is carefully poured into a polyallomer ZO tube in which 2.5 ml of 5.7 M cesium chloride and 0.1 M EDTA
solution (pH 7.5) hay been loaded in advance, and then subjected to ultracent~ifugation at 30,000 rpm for 12 hours at 20 ° C us i ng Becb:man Sta41 r o for (man of a~: t~.:red and sold by Beckman Instrument, U.S.A.). After removal of the supernatant, the pellet is dissolved in 1 ml of 10 mM Tris-HC1 buffer (containing S ~WI EDTA and 1 w/v o SDS) . The resulting solution is extracted with a 4:1 by volume mixture of chloroform and 1-butanol. To the aqueous erase, 0.05 volume of 2.bI sodium acetate and 2.5 volumes of ethanol are added, and allowed to stand at -20°C for 2 hours or more, thereby to precipitate Rt~n. The precipitGt.e is collected by centrifugation, dried, and then cissolved in 500 ~l of sterile water. r.s a result, a cytoplasmic RNA solution is obtained.
The above-obtained RNA solution is heated at 68°G for 2 minutes, and thereaft'r, chilled quickly. 500 ~l of 2-fold concentration 10 m:~i Tris-EDTA buffer (pH 7.4) (containing 1.
1 mM EDTA, 0.1 w/v $ SDS and 0.5p,lithium chloride) is added to '~~' the solution, and the mixture is a lied to a 200 m oli o '" ,' PP g 9 dT-cellulose (manufactured and sold by Bethesda Research Laboratories, Inc., U.S.A.) column, and washed with 10 ml of the same buffer (one-fold concentration) as described above. The material retained by the column is eluted with 2 ml of an elution buffer containing 10 mt~z Tris-HC1 buffer pH 7.4, 1 mM
EDTA and 0.1 w/v ~ SDS. To the eluate, is added 0.05 volume of S ~ ~ ~.~.~'ic ~
sodium acetates and 2.5 volumes of ethanol, and the mixture is cooled at -20oC to precipitat e. The precipitate is collected by centrifigation, and applied to the oligo dT-cellulose column, , and the fractions adsorbed onto the oliao dT-cellulose are collected. 85 ~g of mRNA is recovered as determined by the ultraviolet spectrum analysis.
Step 4 (Size fractionation of mRNA) 880 ~Zg of mRNA prepared by the same method as described in Step 3 is dissolved in 250 ~l of water, and the resulting solution is layered onto a 10 ml 5-25 ~ linear sucrose density gradient. The sucrose density gradient is prepared by means of ISCO 5i0 gradienter (manufactured and sold by ISCO Inc., U.S.A.), using Tris buffer solutions [containing 25 mM Tris-HC1 i (pH 7.2), 2 mM EDTA and 1 w/v ~ SDS] respectively containing 5 sucrose and 25 ~ sucrose.
Using Beckman Sh741 rotor, ultracentrifugation is effected at 40000 rpm for 12 hours at 4oC, and fractions each of 400 ~1 are recovered by means of a fraction recovering apparatus (manu-factured and sold by Beckman Instrument, U:S.A.), and then ethanol precipitated. The precipitated fractions are centri-fuged, and dissolved in sterile water.

1 341 34 $
step 5 (::periment on translation of mR:!=.) Translatio;a of mRW usina oec~.~tes of Xeno~us laevis (Hamamatsu biological teachinc materials) is conducted .according to the procedure described in the experimental reports (~:or exa~ple, HirosPi Teraoka, riikio Itsuki and Kentaro Tanaka, "Protein, Nucleic acid, Enzyme", Genetic Engineering, extra edition. , 1 9S1 , p 602) . Xe_~aopus laevis is procured from Hamamatsu biolocical teaching materials.
Fractionated mRNA obtained in Stec a_ is dissolved in sterile water to have a concentration of 1 ~c;r:l, and the solution is injected into oocytes in such a small amount as 50 nl per cell . Cells are then c~ul t~,~red for 2 "-. hour s in a Bar th' s solution (contai n.i.ng 7. 5 rru~i Tr is-HCl (pH 7. 6 ) , 88 mill NaCl, 1 mrd potassium chloride, 0.33 tnr~! calcium nitrate, 0.41 m--4 cal cium chloride, 0. 82 m_~i a«agnesium sulfate, 2. 4 m:~~ sodium bicarbonate, 18 G/ml pe.~.icil 1 in G and 18 ;.ig/ml s t=eptomvci n]
whi ch contains 1 mg/ml bovine serum al bu:,.in. Ooc,~ tes are crusted, in the c~~lture licuid, by means of a glass bar.
The culture liquid is then centrifuged, and tale supernatant :is evaluated for the cytotoxic activity against L cells.
mRN~ which will be translated to give a polypeptide having maximum activity sediments as 16 S in size. This activity is eliminated by the anti-TNF antibody obtained in Step 3 of Referential E:~ample 2, but is not eliminated by the normal mouse serum.

S tep 6 (Preparation of trans~ormants) Using 5 yc of the fractionates mP,Na obtained in Step 4, a double st~-_~.ndeu D:dfi is prepared in accordance with proce3ure S described in Iterutu_~-e (1) , from pace 96. As the reverse transcriptase, use is made of a product of Life Science, Inc., U.S.A. The double stranded Di~~=~, is size-fractionated on a 3.5 o polyaciylamide gel, and 330 ng fraction of about 1000 to 2000 by is obtained. In accordance with the procedure described in Literature (1), 7 ng of this fraction~~is extended with deo.xyC residues using terminal deoxynucleotidyl transferase (manufactured and sold by .
Bethesda Research Laboratories, Inc., U.S.A.) and annealed with 56 ng of plasmid pBR322 which has been digested with PstI and extended with deoxyG residues. The so-annealed mixture is inserted into E. coli h-12 strain (HB101, ATCC 33694) to transLorm the strain. As a result, 12000 transformants are obtained.
S tep 7 (Partial amino acid sequence of rabbit TNF) Rabbit TNF partially purified in Referential Example 2 (activity: 5 x 107 units) is subjected to SDS-polyacrylamide gel electrophoresis for purification as in Step 2. Part of the gel is dyed with Coomassie Brilliant Blue. A band at the position corresponding to the molecular weight of 17000 is cut out from the gel, and extracted with 1 % ammonium bicarbonate. About 180 ~g of TNF is recovered as pro~ein.
150 ~g cf the recovered TNF is dissolved in 75,~ of 1 ammonium bicarbonate, followed by addition of 3 ~g of TPCK

trypsin (manufactured and sold by Worthington Eiochemical, U.S.A.). The mi:_ture is incubated at 37°C for~4 hours.
c~Il. arc>rmcchc:~
The mixture is then fractionated by means a~~~ J
liquid chromatography column comprising Cosmosil 5C8 (manufactured and sold by Nakarai Chemical, Ltd., Japan) as the packing material, thereby to obtain fragments digested with trypsin.
The highly purified TNF and the trypsin-digested fragments thereof are then subjected to desalting by means of Sephadex G-25 column, and then freeze-dried. According to the method of R.M. Hewick et al (see J. Biol. Chem., Vol. 256, pp 7990-7997, 1981), the purified TNF and the trypsin-digested fragments are each subjected to Edman Degradation from the N-terminal. PTH-amino acid liberated in each step h~gh -~er~tarmccncc~.
is analyzed by t:~ie customary method by means of a.. h.~~-,cr~-chromatography model SP8100 (manufactured and sold by Spectra physics, U.S.A.) using Soloacks ODS~(rnanuyactured and sold by E.I. Du pcnt, U.S.A.) as the column. As a result, it is found that the T_~,F has the following N-terminal a.~,ino acid sequence: Ser-Ala-Ser-Arg-Ala-Leu-Ser-Asp-Lys-Pro-Leu-Ala-His Val Val Vila-?=r.-Pro-Gln-Val -Glu-Gly-Gln-Leu-G1 n-One of the trypsin-digested fragments has the following N-terminal amino acid secuence.
Glu Thr Pro Glu Glu Ala Glu Pro Met Ala ' Step 8 (Synthesis of oligodeoxynulcleotide probe) Oligodeoxynucleotides complementary to the base sequence of the mRNA which is deduceu from the amino acid sequence fir- t~ac~e ma ~' of rabbit TNF obtained in Step 7 of kefarential Example 3 is synthesized according to the i.~,;proved phosphotriester method which has already been reporter by the present inventor in H. Ito et al, "Nucleic Acid Res." 10, 1755-1769 (1982).
In preparing oligodeoxvnueleotides, 128 oligodeoxynucleotides estimated from the amino acid sequence of rabbit T~ZF are classified into five groups, namely groups of 16, 16, 32, 32 and 32 and are synthesized as mixtures of oligodeoxynuc leotides of the respective groups. The obtained oligodeoxy nucleotides of the respective groups are deprotected according to the cus~~~a~r~~m5~ethod and purified by column ~".,~ .J h ,~, -- chromatography using G-50~ (manufactured and sold by Pharmacia Fine Chemicals, Inc., Sc.~eden?, electrophoresis on a 20 b by weight polyacrylamide gel containing 7 hi of urea and column chromatography using DE52~(manufactured and sold by what:nan Ltd., U.S.A.). The thus obtained oligodeo::ynucleotides of the respective groups are dialyzed against 0.1 mri Tris-EDTA buf=er solute=on.
Each or the purified oligodeoxynucleotides of the respective groups is labelled using T4 pol.ynucleotide kinase (manufactured and sold by Bethesda Research Laboratories, Inc., U.S.A.) and ~-32F-adenosine triphosphate according to the customary method and then puri=fied by column chromatography using DE52 (manufactured and sold by whatman Ltd., U.S.A.).
The radioactive material is incorporated into each of oligodeoxynucleotides of the respective groups in an amount '~ 'fi f'~0.~ N PnG~r ~5~

~ 341 34 8 of about 3 x 10g cpm/ug. The oligodeoxynucleotide probes each obtained in the /form of a mixture of the respective group are ' designated as shown in Table 1.
Part of the amino acid seauence of the rabbit TNF, the base sequence of the mRNA estimated from the amino acid sequence of the rabbit TNF and the base sequences of synthetic oligodeoxynucleotide probes of the respective groups are shown in Table 1.
Table 1 .
Amino Carbo::yl Amino acid ~ terminal-.-Ala Met Pro Glu Ala Glu Glu- terminal se~. w:~ce m RNA ~ 3 " - - XCG G XCC YAG XCG YAG YAG- 5' - TA

Probe r~i 5' GC CAT riGGMTC GGC P~1TCMTC3' Probe MI 5' GC CAT NGG MTC GGC biTCMTC3' Probe P~~L7~5' GC CAT ZGG MTC AGC MTC LMTC3' Prone h1K S' GC CAT ZGG MTC CGC AiTCMTC3' Prone ML 5' GC CAT ZGG MTC TGC MTC MTC3' tvote: X represents a ribonucleic acid residue of A,C,G or U.
Y represents a ribonucleic acid residue of A or G.
25 M represents a deoxyribonucleic acid residue of T or C.
N represents a deo~:yribonucleic acid residue of A or G.
Z represents a deoxyribonucleic acid residue of A, C, G or T.

mRNA of the cells producing TNF which is obtained according to Step 3 of leferential Example 3 is treated with mM
a solution containing 1 ~ of glyoxal, 10 of NaH2P04 and a 50 $ by volume dimethyl sulfoxide at 50oC for 60 minutes S and then subjected to fractionation using electrophoresis on a 1.1 $ by weight agarose gel. The fractionated mRNA is transferred on a filter of an electrophoresis type transfer blotting apparatus (manufactured and sold by Bio Rad, U.S.A.) according to the manual of the maker. Then the mRrlA on the filter of the apparatus is. treated with a 5 x Denhardt's solution containing a 5 x SSC solution and 150 ug/ml of denatured salmon spermatozoa DNA at 65oC for two hours and, then treated with a 5 x Denhardt's solution containing 1 x 10~
cpm/ml of the labelled oligodeoxynucleotides and a 5 x SSC
solution at 50oC for two hours. The above-obtained filter is washed with a 6 x SSC solution successively four times at room temperature, 40oC, 50oC and 60oC. An XAR-5 X-ray film (manufactured and sold by Eastman Kodak Company, U.S.A.) is exposed to the radiation from the filter. As a result, it is found that the oligodeoxynucleotides designated by Probe M,1 are most strongly hybridized with the rc,Rr;A, showing that the oligodeoxynucleotide having a base sequence which is completely complimentary_to the mRNA is contained in the oligodeoxynucleotides designated by Probe hiJ.

Step (Cloning of TNF acne orabbit) In accordance with the proc~cure des;_=ibec in Literature (2), page~162, the t:ransformants obtained in Step 6 of Referential Example 3 are transferred onto a cellulose filter and the DNA oz the trans~ormants is h;~bridized with the labelled oligode.oxynucleotide (Probe DIJ) selected in Step 8 of Referential Example 3 under the same conditions as in Step 8 of Referential Example 3 (colony hybridization).
In the just above procedure, 49 colonies which are strongly hybridized with the labelled oligodeoxynucleotides (Probe r1J) are .selected and further fixed onto another nitrocellulose filter. Then, using 49 colonies, further hybridization is carried OLlt. -- -~ ~ -- -t0 Select nr.ne CO~ni?ic~ae ~,-h;C1-; arc ~.,Cra StrC::~:~ly hybridized with the labelled oligodeoxynucleotides (Probe AiJ).
In accordance with the rapid plasmid separating procedure described :in Literature (1), pace 6, about 5~ug plasmid is obtained from each of the nine colonies. Each of the obtained plasmids is cleaved using restriction enzymes, Pstl, TaqI, RsaI and PvuII (each manufactured and sold by Bethesda Research Laboratories, Inc., U.S.A.) according to the procedure described in the manual of tile mai~er, followed by electrophoresis effected on a 1 b by weight agarose gel.
Then, fragments obtained by cleavage by the respective restriction enzymes are compared c~~ith respect to length thereof.
The results suggest that all the nine strains correspond-ing to the nine colonies have the base sequence of the fragment obtained by cleavage by P~.~uII and RsaT_ and consisting of about 50 by and that most or the nine strains ,~~:e the base sequence of the fragment obtained by cle~vace b,; Rs3I
and consisting of about 200 bo. In ocher words, the results suggest that the nine stiains have partially' common base sequences. The results of analysis by the restriction enzymes are shoran ir: Fig . 1.
Seven strains containing plasmids designated in Table 2 below are separately cultivated in 2 ml of LB medium containing 10 ~ g/ml of to trac~Y.cline until the optical ' density of the solutions_shows the values shown in Table 2 below, followzc by centri=ugation to obtain rescective s trains . Each o f the ob to i nod strains i.s separately added into 2 ml of physiological saline and disrupted by sonication.
The obtained solutions are subjected to centrifugation and the cyto toxic ac tivi ty a7air.s t L cells of the obtained SuaernatantS 1S QC'.~arlCt'_:?~'d. Tile r2Si:ltS are ShUwn li Table 2 below. As a blank test, the same procedures as mentioned above are repeated using a strain containing plasmid pBR322.
The results are also shown in Table 2 below.

1341~34g Table 2 I
Number. Cyto to:~ic of ~D600 ~c~ivity annealed against L
base pairs cells Plasmid (unit/ml) pB 2-2 1400 1.369 35 pB 2-3 800 1.605 < 10 pB 2-7 1060 1.364 < 10 pR 9 1550 1.618 < 10 pR 12 1400 1.458 15 pR 18 1850 1.438 < 10 pR 25 1350 1.514 < 10 pBR322 0 1.677 < 10 The cytotoxic activity against L cells is eliminated by anti-TNF antibody but is not eli:r>inated by normal mouse serum. This shows traat all of the above-m~~ationed nine colonies Q have plasmids which contain oligodeohynucl.eoti~s coding for TNF.
Steo 10 (Determination of base sequence of DNA coding for rabbit TNF) E. coli strains containing plasmids pB2-7 and pR l8are cultivated in one liter of M9 medium described in Literature ~ 341 348 (3), page 440 and containing 10 yg/ml of tetracycline. Then, in accordance with procedure described in Literature (3), page 90, each of the plasmids is isolated in an amount of about 150 fig.
The base sequence of the insert of each plasmid is deter-mined according to the riaxam-Gilbert chemical procedure described in Maxam et al "Method in Enzymology", ~, P 490 (1980), Academic Press. The thus determined base sequence is found to be in agreement with the partial amino acid sequences determined in Step 7 of P.eferential Example 3. Thus, the whole sequence of TNF of rabbit is considered to be elucidated.
Step ll In this step, cpnstructica of a plasm,id is carried out using the recombinant plasmid pRl2 to obtain direct expression of TNF in E, c~li using lac as a promoter. The procedures are illustratively shown in Fig. 2. First 10 jig of plasmid pRl2 is digested with 10 units of A~aI (manufactured and sold by Bethesda Research Laboratories, Inc., U.S.A.) at 37oC for two hours and electrophoresed on a 4 ~ by weight polyacrylamide gel to isolate ~~ by fragments. About 1 ~g oi: the fragment is isolated from the gel by electroelution. In the same manner as in Step 8 of Referential Example 3, two oligodeoxynucleotides shown in Fig. 2, namely 5'-GATCCATGTCAGCTTCTCGGGCC-3' and 5'-CGAG~=,GCTG:~C.~-.TG-3' are s;rn:.hesi~ed. Then, e:.ch 5' end of the oligodeo::ynucleotices (about 100 amole) is phosphorylated usinc T4 polynucleotide ~:inase in accordance with the method described in Lterat~~re (3), page 122. After completion of .the reaction, the re.ction mi:~.ture is extracted with phenol and then with chlorofor:,t. Then the obtained synthetic oligomers are mixed with 0.5 pug o. the AaaI
630 by fragment and ethanol precipitated. The fragment is ligated with the synthetic oliaomers at 4°C overnight using 10 units of T4 DNA lipase in accordance with the procedure described in Literature (1), page 37. After comple tion of the raac tion , the rear tion mixtur a i s a thanol precipitated and diaested with 20 units of BamHI at 37°C
for three hours, followed by electrophoresis e=fected on a 4 g by weight polyacyyla~:,ide gel to recover 6 70 by fragment by electroelution. One lag of commercially available plasmid pUC-8 (catalog No. 4°16, manufactursd and sold by P-L Bioche.:,ics,ls, Inc. , U.S.A. ) is digest=ed with Ba:~HI and ex~=acted with phenol and then with chloroform, followed by ethanol precipitation to obtain a vector. 0.~ ug of tine obtainea vector is ligated witz the abov~'-obtained .
fragment having BamHI sites on its both ends and containing about 670 ~p coding for TNF using T4 DNA l.igase. In accordance with the procedure described in Literature (4), page 20, 2 5 ~ . co i l is tr ors for:re~? us ing the above-obtained vector and cultivated on an agar medium containing 1 m_M of IPTG and 0.004 0 (wlv) of X-gal to obtain about 200 c~~otija'~
i *~ . ~'laSiTi? d .iva i5 ~ rcDurO : ~rO.:W.~17 Oi tlleSe transformants and ;'_ices~ed wit: EamcT__ :~s a result, it is found that 15 plan..°.,ids contain the in te:~de~.~ BamHT fragment (about 670 bp) . I- order to e:~a:r4i ne the correction of insertion, the abo-.~e 15 alas:a:ids are digested with EcoRI
having only one reYognition site on its pUC-8 and PvuII having only one recognition site on its about 670 base pair fragment part and electrophcresed on a 6 ~ by weight polyacrylamide gel : As a resent, i t is determined tha t 7 plasmids have the intended fracr.:e.~.t consisting of about 140 bo and that the direction of transcription o' the lac promotor on pUC-8 is in agreement with that of the olicodeoxvnucleotides coding for TNF.
D:3A seauence ':gal ysis shows tha t these seven plasmids have the same seeue:-:c' and have the desired nucleotide sequence at the ] ua~.C tlOnS '1..'e-..wiee:: tile lac promo ter , SV.~..''.a'1°t1C D\pa and C~iJ.~a.
-=; -, - o p ; o.: using Cons~_u..~_Ln cr fur ~h-r las.:~ids is carr_~.,. ouc the recombinant plGS:;id pRl7 in order to obtain direct e::pression of TVF i:: ; coli using lac UV5 as a promoter.
The procecures are =.llustratively s:~o~cn in Fig. 3. First, 10 ug o= the nlas;.i..d pRl i is dices ted with 10 units of Anal (manufacture: and s:.ld by Bethesda Research Laboratories, Inc, U.S.y.) at 37°C for tc,;o hours and elec;rophora_sec on a 4 0 b~' wee gzt polyacryl_::ide gel to isolate a fragment consisting of about '030 b~:. rbout 1 ug of tie rragment is isolated from the gel by ele~~roelution. In the same manner as in Step 8, two oliaodeo~:,rnucleotides shown in Fig. ?, namely S'-~.yTTCATGTCAGCTTCTCvvGCC-3' and 5'-CG~Grl.=;GCTGACATG-3' are synthesized. Then, each 5' end of the two oligoc.eo wnucleotides (about 100 pmole) is phosohorylated using T~ polynucleatide ;:inase in accordance with the method described in Literature (3), page 122. After completion of the reaction, the reaction mixture is e~:tracted with phenol and then with chloroform. Then th.e synthetic oligomers are mixed with 0.5 ug of the previously obtained A~aI
fragment (about 630 bp) prepared from the plasmid pRl7 and ethanol precipitated. The fragment is ligatec with the synthetic oligomers at ~°C overnight using 10 units of T, liaase in accords.~ce 4;=th pyocevure deso.ribed in Literature (1), pace 37. After co~:cletion of the reaction, the reaction mixture is ethanol preciti rated and digested caith 20 units o~ EcoRI at 37°C for threw hours, fallocaed by electrophoresis a~fected on a 4 o by weight polyacrylamide gel t.o recover a frag~~,e:.t (about 670 bp) by electroelution.
In accor6ance with the procedure described in F. Fuller, "Gene", 19, pp 4?-54 (1962), plasmid pOP95-15 is prepared.
Gne ug of pOP95-1~ is digested with EcoRI and extracted with phenol and then with chloroform, followed by ethanol precipitation to obtain a vector. Using T~ DNA ligase, 0.5 ~g oL the obtaine& vector is ligated with the fragment z5 (about 670 bp) obtained by ligating the synthetic oligonucleotide with the oiigonucleotide cooing for TNF.
In accordance with the procedure described in Literature (4), - sa -1 34i 348 page 20, E. coli Ji~I101 (ATCC 33876) is trans=ormed using the above-obtained vector and cultivated on a medium containing 1 m'~I of _TPTG and 0.00 ° (sa/v) o:: X-gal to obtain about 150 tchite colonies. Plasmic DIZA i s prepared from 100 of these colonies and digested with EcoRI. As a result, it is found that 12 plasmids contain the intended EcoRI
fragment (about 670 bp). In order to ewamine the direction of insertion, the above 12 plasmids are digested with PvuII and PstI and electrophoreses on a 1.5 ~ by weight agarose gel. As a result, it is determined that our plasmids havethe desired fragments (about 1280 by and about 2600 bp) and that the direction of transcription of tine lac, UVS
prc,-.~ot.er is in agreement with that of th~.e oligodeoxynucleotides coding for TNF. .
Ease secuence analysis shows that these four plas,«ids have the same seguence and that the lac UVS promoter, the s:~.~.tze tic oli godeoxyrucleotide and cD~lA are prooerl y co:~.bined with each other. The obtained plasmids are designated pTIF-lacUVS-1.
Stew 12 (PUri.fication of TNF produced by E. coli) E. coli strains containing plas;~ids obtained in Step 11 are cultivated e~rnr 50 m1 of LB medium containing ~9.~g~.u~.
~- ampicillin at 37°C overnight. Then the strains are transferred to 5 liter of LB medium containing 100 ugjml of ampicillin and further cultivated at 37°,C for thrlee hours.

1 341 34$
Tsopropyl- s -D-thiogalactopyranoside (manufactured and sold by Sigma Chemical Company, Inc., U.S.A.) is .added to it to a final concentration of 1 mM. Further cultivation is carried out for six hours, followed by cooling. Then str<~ins are collected by S cent,~ifugation. In the saruc manner as described in Step 11, the ,, ' r a~O S
,~~, are added into 5 liters of 0.04 M Tris-HC1 buffer solution (pH 7.8) and disrupted by sonication to obtain a strain protein solution. The obtained solution has cytotoxic activity 7 ~
against L cells cf 5 x 10 tmi.
The obtained solution is purified in the same manner as in Step 2 of Referential Example 2 to obtain 1.2 x 106 units of TNF. The specific activity of the TPJF is 6.8 x 10~ units/mg.
Step 13 (Eva2uation using transplanted rieth A
sarcoma in mouse) -2 x 105 rieth A Sarcoma cells are transplanted intradermally in the abdominal area of a BALB/c mouse and, 7 days laterr mice with tumors of 7 to 8 mm in diameter and with no spontaneous central necrosis are selected for evaluation. A sample (0.2 ml) of TNF obtained in S~ep 12 of Ref erential Example 3 and diluted with physiological saline solution is injected through the tail vein. The activity of the sample is - 5z -' 1 341 348 evaluated after 24 hours accordinc to the follo4:inc criterion.
(-): no change (+):~ slight hemc=rhagic ne:=osis (++): moderate hemorrhagic necrosis (central necrosis extendinc over approximately 70 'o of the tumor surface) (+++): marked hemorrhagic necrosis (massive necrosis leaving a small viable rim along the tumor periphery) 20 Days after the injection of the sample, observations are made on the involution of tumors and recovery rate is determined according to the follocaing equation.
Number of 'mice which had been completely recovered from tumor Recovery rate = _ _ Number o~ mice used nor test The results are mown in Table 3.
Table 3 Injected amount Evaluation for _ of rabbit Ti~'F Number of activity of Recovery produced by mice used samples rate E. coli for test (af ter 1 'day) (after 20 days) units/mouse - + ++ +++

2 x 105 5 0 0 ~ -4 5~S_-Reference 5 5 0 0 0 0/5 (physiological swine ) Example 1 Step 1 (Trans'ormation of eoli Fl? Strain riC1061 cvith -pRl3, pa2_7 and pE2-2 ?lasmids) Colonies of E. coli K12 strain tIClOo'1 are transformed with eac:~ of the pRlB, p52-7 and pot-2 blasmids, which are obtained in Reference Example 3, according to the customary procedures. Specifically, colonies of E. coli. K12 strain MC10~1 are cultured in LB medium until the or~tieal density of the cul tore broth beco~s 0 . 3 at 550nm. 50m1 of the grown E. coli cult~.:re is harvestc~, washed with a 25m1 mix-ture containing lOmM b?OPS(pH7.0) and lOmM RbCI, and re-suspended in a 25m1 yixture containing 0.1M MOPSraH6.5) , 50mM
Carl and 1 On,M obVl . T!~c resu' ti:.a su:~oen sion is cooled on ice for 30 min ,.centriTuced and suspended in a mixture of 2ml eT the above-::Ze.~.ticnec mixtLre COntai:~ing O.1M
P~~OpS (0:6.5) , 50mf~? CaCl2 and 10:«.x'. P~bCl and 30 ~1 of D_~:50.
To a 200 ~l alicuot of the resulting suspension is separately _ _ _ acded 10 ~:1 of each of the plas:aid DVfi solutions. Each of the resulting mixtures is cooled on ice for 30 min, and glen h2at-SIlOCVe.~.. at $a°C fOr 6~ SeCOndS. In'uT,edlately thereafter, 5Ti1 of the LB medium pre-warmed at 37°C is added to each o.-'. t:7e heated mixtures, followed by incubation at 3'~' ° C f or one hour . The ob tame cui tyre bro the are each subjec~e~ t0 C2ntrifugation to form cell pellets. The super-nata.nt is discarded, and L3 medium is added and stirred to resuspend each of the cell pellets. Each o.f the resulting suspensions is inoculated to an Ln agar ;late containing 30 ug/ml tetracycline, followed by incubation at 37eC
overnight. As a result, colonies of tetracycline-resistant transformants transformed, each, with pRlB, p82-7 and pB2-2 plasmids are obtained.
Step 2 (Preparation of pB2-7 and pRl8 Plasmid DNAs) Each of the trans=ormants respectively transformed with p82-7 and pF,l8 plasmids which are obtained in Step 1 is subjected to (1) growth of the transformant and ampli-fication of the plas;nid; (2) h=_rvesting and lysis of the transformant, and (3) purification of the ;plasmid DNA, in accordance with the procedures as desc=ibed at Dages 88-96 of T. h~aniatis, E. F. Fritsch and J. Sambrook, "Molecular Cloning", publ_shed by Cold Spring Harbor Laboratory, U.S.A.
Illu~trativelv stats~ a r =ormants is inoculated a ch of th t_Gns~
x ~,. c~r~ taa~r ~~~~~~~m ~ -f-efirc~ c~.~c/;ne- _ into LB medium ~ and ~.nc ated at 37 C with vigorous ~a_ shaking. This step is repeated to attain growth of the transformant and amplification of the plasmid. The tansform~nt cul tune is harvested by centrifugation at 40008 for 10 min.
at 4°C. The sub... ata.~xt is discarded. The resulting pellet is washed in 100 ml of ice-cold STS [O.1M NaCI, lOmM
Tris~C1(pH7.8)/ a'nd lm:'~I EDTn], and subjected to lysis by )Y'7 Q SO~~c-~'lOn C,~ _ boiling by-~3~e-~of 20 mg/ml lysozyme 10 mM Tris-Cl, pH 8Ø
The viscous product is transferred to an ultracentrifuge tube, and centri~uged at 25,000 rpm for 30 min at 4°C to obtain a DNA solution. The volume of the DNA solution is measured. For every milliliter, elactly lg of solid cesium chloride is added and mixed gently u:.til all of the salt is dissolved. 0.8 ml of a solution of ethidium bromide (lOmg/ml in H20) is added for every lOml of cesium chloride solution. The final density of the solution is 1.55g/ml, and the concentration of ethidium bromide is approximately 600 ug/m1. The cesium chlor:i.de solution is transferred to a tube suitable for centrifugation, and the remainder of the tube is filled with light paraffin oil.
Centrifugation is conducted at 45,000 rpm for 36 hours at 20°C to obtain two bands of DNA, the upper band thereof consisting of linear bacterial DNA and nicked circular plasmid DNA and the lower band thereof consisting of closed circular 1~ plasmid DNA. The lower band of DNA is collected into a glass tube through a hypodermic needle inserted into the side of the tube. The ethidium bromide is removed, and the aqueous phase is Dialyzed against TAE. The plasmid DNA
solution is treated with RNase, and extracted with an eaual volume of equilibrat~3 phenol. The aqueous phase is layered on a column of Bio-Gel A-150 equilibrated in TAE
(pH8.0) and 0.1 $ SDS. The DNA in the column is washed, and a reservoir of TE with 0.1 °a SDS is applied to collect fractions. The fractions are precipitated with ethanol to obtain a pure plasmid DNA, *trade-mark By conducting the above procedures, 250 jig of pure pB2-7 plasmid DNA and 134 ~g of pure pRlB plasmid DL3A are obtained.
Step 3 (Nick Translation of Pure pB2-7 and pRlB
S Plasmid DNAs) From the pure pB2-7 plasmid DNA obtained in Step 2, 40 ~g is taken, digested with PstI restriction enzyme and subjected to electrophoresis through 4 $ acrylamide gel. After electrophore-sis, the DNA is stained and the desired band is cut out to isolate a Pstl insert.
Using 500 ng of the isolated Pstl insert, nick translation is carried out in the manner as described in Maniatis, T, et al, proc. Natl. Acad. Sci. U.S.A., 2.2, 1184 (1975). For the nick translation, the Nick Translation Kit produced and sold by Bethesda Research Laboratories Inc., U.S.P,. is employed, and 80 pmole of radioactive dCTP is applied in a 25 ~1 reaction system (at 400 Ci/rr~mole) . To a mixture consisting af:
2.5 ~1 Solution A (dN'lyP's solution) 2.5 ~1 Solution B (500 ng of test DNA viz. PstI insert) 5 ~1 hot dCTP (3200 Ci/mmole) s D p,~Je~la 1 dc~r.p =n, 1.3 ul cold dCTP (65 pmole, Solution E (Fi2p) 22 .5 ul (total ) is added 2.5 ul of Solution C (DhaseI, Div:y Polvmerase I), and reacted at 15°C for 60 min. Then, Solution D (stop buffer) is added to the resulting mi::w.:re to stop the re-action. Further, carrier tRNA is addec, subjected to ethanol precipitation twice and dissolved in 500 N1 of water.
The specific activity per ug DNA is 9.3 x 107 cpm.
With respect to the pure pRl8 plasmid DNA obtained in Step 2, also, the above-described procedures are carried out to effect the nick translation. The saecific activity lS _ per ug DNA ~s 7 x 107 cpm.
S tep 4 (Preparation of RsaI Insert E'ragment of pRl8'Flasmid DNA) 80 ug of the pRl8 plasmid DNA is digested with RsaI
restriction. enzyme, and subjected to electrophoresis through 4 % poiyacrylamide gel. The following ces:~r~ i~.:.n.~'.s of ins~~s are gist out a~ purified by means of the E:D column about 640 by 3.77 ~g (recovery 52. %) about 175 by 1.77 ~tg (recovery 50~ ~) .
The above about 640 by insert is designated as 3'-frarm~nt of pRl8 (meaning 3'-untranslated region or pRlB), and the above about 1 i 5 by inset t is designated as pRl8-cfr (r~:.,ning coding region of pRl.B) .
1~"~reover, the above procedures are . repeated using PstI and MStII
res tr is tion enzymes istead of the FsaI res trio tion ~enzvrr~ to obtain the following band:
about 450 by 3.65 ~Zg (recovery 60go) The a~ve inse_~-t i s designated as 5' -fragment of pFtl8 .

Step 5 (Isolation of the Human Genomic TNF Gene) The 32P-labelled plasmid pB2-7 insert obtained in Step 3 of Example 1 is used as a hybridization probe: to screen 106 plaques of bacteriophage Charon 4A/human genomic lihrary prepared by insertion into the Charon 4A EcoRI ligation site [Elattner et al, "Science" 1~, 161 (1977)] of sized fragments from partially digested human DPdA [Maniatis et al. "Cell"' 3..5., 687 (1978) ] . The plaque hybridization method of Benton and Davis [Benton and ZO Davis, "Science", lq.f~, 180 (1977)] is used. Since not all of the bacteriophage in the starting culture contain the necessary genetic material for preparing human TNF, a probe which has a bash seque::ce co:pl eme, tart' to the La.~"'.rblt 'f~:F gene is used. DiaA
of phage plagues having the desired genetic material incorporated the radioactive probe and are identified by their radioactivity. Nine hvbricizing plagues are isolated from the library.
The proceaures and conditions used are as follows.
1) Number of plaques:
ZO ~-1 x 106 plaques (~ 4 x 104 plaques/~5150 mm plate x 25) 2) Transfer to nitrocellulose filters:
[see Benton ar_d Davis, Science, 13.x, 18G (1977) ) 3) Hybridization:
Addition of 1.25 x 105 cpm/ml of pB2-7 insert probe prepared in Step 3 of Example 1, 42oC, 19.5 hr 4) Washing:
2 x SSC - 0.1 % SDS at room temp.
Immersion'10 min. x 4 1 x SSC -~~0.1 % SDS at 50oC
Immersion ~0 min. x 2 5) E: posure:
XAR-5 (Eastman Kodak Company, U.S.A.) -80oCr 2 intensifying screens, 39 hr In the above screeningr 12 candidate strains are obtained.
In the same manner as mentioned above, second screening is carried out to obtain nine strains containing the intended fragment. Using these strains, third scrs~ening is carried out in the same manner as mentioned above to obtain nine strains containing the intended fragment. Using t_he obtained strains,' fourth screening is carried out to confirnn that the nine strains contain the intended fragment. The obtained nine bacteriophages containing the intended fragment are designated HG-1~~ HG-9, respectively.

Step 6 (Isolation of Rabbit Genomic TNF Gene) Substantially the same procedure as described in Step 5 of Example 1 are repeated except that 106 plaques of bacteriophage Charon 4 A/rabbit genomic library which is prepared using digested rabbit DNA [Maniatis et al, Cell, 1.~. 687 (1978)) instead of digested human DNA. 6.7 x 105 plaques of bacteriophage Charon 4A/rabbit genomic library are used instead of 106 plaques of the bacteriophage Charon 4A/human genomic library. Thus, there is obtained two bacteriophage strains (RG-1 and RG-2) containing the rabbit genomic TNF gene.
Steu 7 (Southern blotting analysis of human clones) Using the bacteriophages HG-3, HG-6 and HG-7 obtained in Step 5 of Example 1, DNA of each bacteriophage is obtained according to the following procedures.
6 x 1010 cells of E. coli LE392 (host cell) are suspended in 18 ml of SM and 3 x 109 PFU of bacteriophage HG-3 is added, thus allowing the E. ~oli to be infected at 37oC for 20 minutes.
Ther., the obtained mixture is added in 3 liters of NZ-broth and subjected to shaking culture at 37oC for 23 hours. 60 ml of CHC13 is added to the mixture and further subjected to shaking culture for 30 minutes. ~~' NaCI is added to the mixture to a final concentratiorx of 1 M, the mixture is allowed to stand f or 15 minutes, followed by centrifugation to obtain supernatant. Then, polyethylene glycol (molecular weight: about 6000) is added to the mixture so that the concentration of polyethylene glycol becomes 10 ~ (w/v), and allowed to stand for 22 hours at 4oC. Bacteriophages are collected by centrifugation. The obtained bacteriophages are suspended in 28 ml of Sri and an equal volume of CHC13 is added. After stirring by means of Vortex for 30 seconds, the mixture is subjected to centrifugation to obtain aqueous phase. 5M is added to the aqueous phase so that the total amount becomes 30 nl. 26.4 g of CsCI is added to the obtained mixture and dissolved gently, followed by ultracentrifugation (45000 rpm, 20 hours) to obtain bacteriophages in the form of a band. The obtained mixture containing bacteriophages is dialyzed against 10 mM NaCI - 50 mM
Tris (pH8) - 10 mM h:gCl2. Then, EDTA, Proteinase K and SDS are added to the mixture so that the concentrations of them are 20 mM, 50 ~g/ml and 0.5 ~ (w/v), respectively. Then the mixture is treated at 65oC tar one hour and extracted with phenol, a mi:aure of phenol and CHC13 (1:1 by volume) and then with CHC13.
The obtained aqueous phase is dialyzed against 10 mM Tris (pH8) - 1 mM EDTA. The ultraviolet absorption measurement of the obtained aqueous phase shows that pure DNA of the bacteriophage HG-3 is obtained.
Substantially the same procedures as described with respect to the preparation of DNA of the bacteriophage HG-3 are repeated to obtain DNAs of bacteriophages HG-6 and HG-7.

Thus,there are obtained 2920 ~c of HG-3, 1100 ~g oL HG-6 and 819 y~g of HG-7.
In accordance with the Southern method [E.t~i. Southern, J.~lol.Biol., 98, 503 (1975)), Southern blotting analysis of the obtained DNAs is per~ormed. The procedures and conditions are as follows.
1) DNA:
HG-3 825 ng eac:h HG-6 935 ng eacih HG-7 685 ng eac:h 2) Digestion with various restriction enzvmes:
10 units BamHI, 10 units EcoRI, 10 units BamHI + 10 units EcoRI
10 units ni:~dIII , 10 units HindIII + 10 units EcoRI
10 units PvuII
37°C, 3 hr 3) Electrochoresis:
0.8 o Agarose gel TAE
2B V, 15.5 hr 4) Trans~~=_r to nitrocellulose filters:
(see E.ri. Southern, J.r4ol.Biol., 98, 503 (1975)]
5) Pre-hybridization:
ml FDSS
42°C, 6 hr 6) Hybridization 5' - fragment (1 :. 105 ccm/m1) of pRl8 (prepared in Ste:~ a o E:; ample 1) 4"'°C, 14 hr ?) Washing:
2 x SSC - 0.1 o SDS at room temp.
Immersion 10 min. x 4 1 x SSC - 0.1 o SDS at 50°C
Immersion 30 min. x 2 8 ) E~:posure :BAR-5(Eastman Kodak Company, U.S.A.) --8C°C, 2 intensifying screens, la hr The results of hybridization are shown in Table 4.

Table 4 Hybridizing E fragment size with Probe Cl nzyme one oRl8 _ (lxc~e_~io-5 ' end 3 ' end I
pnaQe>

H~-3 6:7 kb BamHI -6 11..2 kb -7 I 9 . 2 kb BamHI HG-3 ~ 2.9 kb +

EcoRI -7 "

EcoP,I -6 "
~ ~. i -7 " r . I ~- I

HindIII HG-3 " ~ j i + _ 6 ~~ , EcoRI -7 ~ " i , i I
HG-3 ~ 9.7 kb 1, , I
~ I
' Hi:~dIII -6 i i i 4 1 kb -7 ~ 9.7 kb ' I
HG-3 ~2.2 0.9 kb kb PvuII -6 1.9 0.9 kb kb -7 ~ 2.2 0.9 kb i ~:b NOTE: The symbol. " ~ " means same fragment hybridizes.

~ 341 34 8 Step 8 (Southern blotting enalysis of rabbit clones) Substantially the same procecures a.s in Step ? of Example 1 are repeated except that each of the bacterio-phages RG-1 and RG-2 is used instead of each of the bacteriophaaes HG-3, HG-6 and HG-?. Thus, there is performed Southern blotting analysis. As a result, it is found that pRl8 5'-fragment is hybridized with a single band fragment of fragments which are obtained by cleavage of RG-1 and RG-2 with each o:~ BamHI, EcoR2, BBC lII, HindIII and BamHI + EcoRI .

', Step 9 (Construction of bacterial clones containing human genomic TNF gene) The method of Landy et al [Biochemistry, Vol. 13, 2134 (1974)] is used to obtain DNA of HG-3 as obtained in the above Step 5. 33 ug of the resulting HG-3 DNA is digested with 80 units of EcoRI at 37°C for 3 hours. The digest is electrophoresed on 1 % low melting agarose gel (conditions:
1 x TAE, 20 V, 14.5 hr). The 2.9 kb band :is isolated from the agarose gel as described by T. Maniatis [Molecular Cloning, Cold Spring Harbor Laboratory, p :377 (1982)].
Specifically, the cut-out gel of the 2.9 kb band portion is heated at 65° for :15 min. The EcoRI-cleaved HG-3 fragment having a length of 2.9 kb (hereinafter often referred to as "HG-3/EcoRI 2.9 kb fragment") is recovered from the melted gel by extracting :3 times with phenol and then 3 times with another extraction solvent, followed by precipitation with ethanol containing ammonium acetate. Thus, there is obtained 637 ng (yield: about 30 %) of HG-3/EcoRI 2.9 kb fragment.
255 ng of the above-obtained fragment is ligated to 56.5 ng of EcoRI-cleaved pUC 13 [J. Messing, Methods in Enzymology, Vol. 101, 20 (1983)] using 2.5 units of T4 ligase at 4°C for 20 hours.
E. coli K 12 strain JM83 is transformed using the above-obtained ligation product. Specifically, E. coli K12 strain JM83 is cultured in LB medium until the optical. density of the culture broth becomes 0.3 at 550 nm. 5G ml of the grown E. coli K12 strain JM$3 culture is collected, washed with ~ 341 348 a ~5 ml of 10 m:~i MOPS(pH7.0)-10 m~! R~C~, and resusae~dec into a 25 ml of 0 . 1 M iwlOPS (pH6. 5 ) -50 m:,! CaC? 2-10 m_~i PbCl .
The suspension is cooled on ice ~er 30 min., centrifuged and resuspended in a mi::ture of 2 ml of 0.1 r1 MOPS (pH6.5)-SO mM CaCl2-10 mrl RbCl and 30 ~1 of Dt~:SO. To 203 ~1 of the suspension is added 10 ~1 of an aqueous lication product solution containing 10 ng of the ligation product. The mixture is cooled on ice for 30 min. and then heated at 40°C for 60 seconds. Immediately thereafter, 5 ml of LB broth pre-warmed at 37°C is added to the heated mixture, followed by incubGtion at 37°C for one hour. The obtained culture broth is subjected to centrifugation and the supernatant is removed.
A.~. Ln ::;~"_;::a is adced to the resultina cell pellet and then inoculated on an LB plate containina 30 ug/ml ar,.picillin and 40 Yg/ml ?~-gal. Colonies containing E. coli K12 strain JM83 which have been transformed with the plasmids having the insert are white, while those containing E:. coli K12 strain JM83 which have been transformed with plasmid only are byte. The obtained white colonies are inoculated again on LB plate containinc 30 ug/ml. a~~icillin and 40 ug/ml X-gal for the purpose of conf i r:,ia Lion .
From the above-obtained white colonies ten colonies (bac~erial clones) are selected and screened by using a mini-prep technique.
Specifically, each colony is cultured overnight in LB medium containing 30 ugjml ampicillin. The grown cells .t,J,~~'°r~ c:c:~n~air~i~e~
are collected and suspended in ~ '~ 2 mg/ml lysozyme-50 mM
,~~
glucose-10 ;~u~1 EDT?-25 m~I Tr i s i~~: 1 (pie . 0 ) . The suspension is allowed .to stand at room temoerat;~re for 5 :,ii:~utes, followed by addition of 200 ul of 0.2 N i~apH-1°s SDS. fter slowly stirring, the suspension is allowed to stand at room temperature for .~. min. Thereafter, 150 ~.:,1 of 3 M sodium acetate (pH5.2) is added, allowed to stand at -20°C for min., followed by centrifugation for 15 min. to recover the resulting suz~ernatant. To the supernatant is added ~1 10 '900 ~- of cold ethanol, followed by centrirugation for 5 min.
to obtain the resulting precipitate. Th~~ obtained precipitate Washed with 70 ~ ethanol and dried to aet a plasmid DNA.
In the above-mentioned method, ten plasmid Dt~As are obtained.
Each olasmid DNA is dissolved in 10 True! Tr i s-0.1 mM
15 F.D1.-~,(~tI~.O), ClCeSteC1 Wlt~'1 yCORI anG'~ sub~eCte.~., t0 ele.~.trOpnOreSlS
for res~r=~tion analysis. The c:,ndit'_ons for cigestion and electroc:zoresis are as follows.
Di~es:._on: plas:~id DNA solution, one-fif th of the a:,iount as p reparea above ; ~ co~I , 3 uni is ;
37°C; 1.5 hr trl eW. j agGr~'r~"e eel; 1 X 1~ ; 1 ro~hcresis : 1 °~ '~ ~ 120 V-2 hr The above restri~~ion analysis shocas that eight of ten clones are pos.tive. That is, the eight clones have 2.9 kb fragment. From tine eight positive clones one clone is selected and cesignated as E. coli ::1.2 strain Jl~i 83 (phGc,) (ABC 3965c Substantially the same proceures as in the above Step 2 are repeater to prepare 1.29 mg of,pHGE DNA, except that E. coli F;12 strain JN!83 (oHGE) is used instead of E. coli harborinj p~32-,' and pF~lS.
Step 10 (Subcloning of EcoRI-cleaved RG-1) 30 ug of RG-1 as prepared in the above Step 6 is digested with EcoRT. From the. resulting fragm~.~nt mixture the fragment havinc; a length of about 3.t~ kb is recovered in substantially the same manner as in the above step 9, except that the above prepared iracment rnixture and d.8 0 lo~,~ melting agarose gel ara cased. There is obtained 1.0 ug of EcoRI-cleaved RG-1 fragment (about 3.5 kb).
The above-obtained EcoRI-cleaved RG-1 fracment (3.5 kb) is 1=~a_o.7 L,J r..,..,~7T_~ ~.~'e,t'y ~..:i'13-. .. .n S'.:~.")~~v..ai~tially the Sable r"anner as in the above step 9, except that the above-obtained EcoRI-cleaved fragment (3.5 kb) is used instead of EcoRI-i ~.-. A,: uG-3 =Y.1. TP ( Q ) .
C_~..v_~ :._..a...~nt 2. kb The transfor:aation of E. coli K12 strain J:~I83, screening o- bac:e=ial clones, dicestion of clones and electrophoresis are. e=fec~2d in subs~a:.tially the same manner as in the above 2 0 Step 9, e_vcept i~hat the a~ve-obtained ligati on product s s- used < The obt.--_~.n ~ clone is aesia:~at w as E coli I~12 s'-~..rai..n J'i~.'83 (pRGE) (ATCC 39655) .
Substantially the same proceaures as in the above St°p 2 are repeated to prepare 1.70 mg of pRGE DNA, except that E. col i hi.2 strain ,T~_33 (pRGE) is~ used ins~ta='d or pB2-7 and pR-18.
Stew 11 (Restriction enzyme analysis of pHGE plasmid DNA) The restriction enzyme analysis of pEiGE DNA as obtained in the above Step 9 is effected accordin.c to the met:~od as described in hlaniatis [~?olecula~ Cloning, Cold Spring Harbor Labora tory, 9e (1982)].

The proc edures and corcitions used are as follows.

1) Diae::tion o= pHGE DNA with EcoRI:

18.5 ~,:g pHGE DNA

64 units EcoRI

37C, 2 hr 2) Ethanol precipitation: precipitate 3) Addition of distilled water to precipitate:

Preparation of 1 uc/~1 EcoRI-cleaved pHGE soln.

4) Digestion with various restriction enzymes:

L:C~ =~i~E~'~~=I
t Restriction enzyme: 5 u-:its PvuII, 5 units Pvu,II f 10 units RsaI, 10 units Rsal, 4 units P2stII, 3 units AvaI, 9 uni is PstI

37C, 2 hr 5) Electrophoresis:

2 ~ Agarose gel, 1 x T'.E, 28V, 14.5 hr 6) Trans=er to nitrocellulose filter:

[see E.M. Southern, ,3. r?o1. Biol., 8,503 (1975)]

7) First pre-hybridization:

30 ml FDSS

42C, 6 hr 8) First hybridization:

5'-fragment (5 x 104 cpmjml) of pRl8 tprepared in the above Step ~) 42°C, 14 hr 9 ) ~,Tashina 2 .; SSC - 0.1 $ SDS at room temp.
S Immersion 10 min. x 4 St~S
1 x SSC - 0.1 °s ~ at 50°C
Immersion 30 min x 2 10) Exposure:
XAR-5(Eastman Kodak Company, U.S.A.), -80°C, 2 intensifying screens, 17.5 hrs 11 ) Washi.ng cut 0 . 5 N! IvTaOH - 1. 5 M haCl ( Immer sion : 1 min . ) 0.5 M Tris - 1.5 M NaCI (Immersion: 1 min.) 3 x SSC (Immersion: 1 min.) 12) Exuosure:
E~~ectea is the same manner as in the above 10), except that exposure time i.s 1° hrs.

13) Second pre-hybridization:
In the same manner as in the above 7) 14) Seconc: hybridization:
p82 - 7 insert (prepared in the above Step 3), 42°C, 16.5 hrs 15) Washing:
In the. same manner as in the: above 9) 15) Eaoosure:
In the same manner as in the above 10), except that e:~:posure time is 19.5 hrs.
17 ) j,iashing out In the same manner as in the above 11) 18) Exposure:
In the same manner as in the above 10), except that exposure time is 20 hrs.
19) Third pre-hybridization:
In the same manner as in the above 7).
20) Third hybridisation:
3' - fragrnent(4.5 x 105 cpm/ml) of pRl8 (prepared in the above Step 4), 42°C, 15 hr.
21) Washing:
In the same manner as in the above 9).
22) Exposure:
In the same manner as in the above 10).
The resul is of the res tric tion a nzyme analysis are sh own in Fig. a.
Step 12 (Restriction enzyme analysis of pRGE plasmid DNA) In subs~antially the same manner as in the a:.bove Step 11, the restriction enzyme analysis of pRGE plasmid DNA prepared in the above Step 10 is effected, except that pRGE plasmid DNA is used instead of pHGE plasmid DNA. The restriction - 73 _ '1 341 34 8 map of pRG~ DNA inse~:. obtaine~ is s:ZOwn in Fig. 5.
Step 13 !Determination of base seauences of rabbit TNF gene and human TNF gene) Substantially the same procedures as in the above Step 2 are repeated, except that E. coli K12 s~rain JM83 (pHGE) obtained in the above Step 9 and E. coli K12 strain JM83 (pRGE) obtained in the above Step 10 are used instead of E. coli K12 strain MC1061 having pB2-7 and E. coli K12 strain MC1061 having pRlB. Thus, 150 yg of each of pRGE
plasmid DNA and pHGE plasmid DNA is obtained.
- The base sequences of pRGE and pi-iGE are determined according to the riaxam-Gilbert method [Maxam et al, Methods in Enzymology, Vol. 55, 490 (1980) published by Academic Press] .
The base sequence of pR-18 determined. in Re=erential Example 3 is compared with that of pRGE as determined above to elucidate the structure, including exon and intron, of rabbit TNF gene. The structure of pRGE DNA insert is shown in Fig. 5. Subsequently, the base seauence of pnGE
is compared with that of pHGE to investigate the homology and consensus sequence around the boundary between intron and exon. Thus, the structure, including exon and intron, of human TNF gene is elucidated. The structure of human TNF gene is shown in Fig. 4.

The above-obtained base segue.nce coding for rabbit T~tF
and human TNF will be shown below. In the base seqnences, the upper~row shows the base sequence Going for rabbit TNF
(R) and the lower roe: the base seauence cooing for human TIvF (H) .
R TCA GCT TCT CGG GCC CTG AGT G?C AAG CC'T CTA GCC CAC GTA GTA
H TCA TCT TCT CGA ACC CCG AGT G~C AAG CC'T GTA GCC CAT GTT GTA
R GCA AAC CCG CAA GTG GAG GGC CAG CTC CAG TGG CTG AGC CAG CGT
H GCA AAC CCT CAA GCT GAG GGG CAG CTC CAG TGG GTG AAC CGC CGG
R GCG AAC GCC CTG CTG CGC A.~1C GGC ATG AAG CTC ACG GAC AAC CAG
H GCC AAT GCC CTC CTG GCC AAT GGC GTG GAG CTG AGA GAT AAC CAG
R CTG GTG GTG CCG GCC GAC GGG CTG TAC CTC: ATC TAC TCC CAG GTT
H CTG GTG GTG CCA TCA GAG GEC CTG TAC CTC ATC TAC TCC CAG GTC
R CTC TTC AGC GGT CP.A GGC TGC CGC TCC ~~~ TAC GTG CTC CTC ACT
H CTC TTC AAG GGC CAA GGC TGC CCC TCC ACC CAT GTG CTC CTC ACC
R CAC ACT GTC AGC CGC TTC GCC GTC TCC TAC CCG AAC AAG GTC AAC
H CAC ACC ATC AGC CGC ATC GCC GTC TCC TAC CAG ACC P.AG GTC AAC
R CTC CTC TCT GCC ATC AAG AGC CCC TGC CAC CGG GAG ACC CCC GAG
H CTC CTC TCT GCC ATC AAG AGC CCC TGC CAG'~ AGG GAG ACC CCA GAG
R GAG GCT GAG CCC ATG GCC TGG TAC GAG CCC ATC TAC CTG GGC GGC
H GGG GCT GAG GCC AAG CCC TGG TAT GhG CCC ATC TAT CTG GGA GGG
R GTC TTC CAG TTG GAG AAG GGT GAC CGG CTC AGC ACC GAG GTC AAC
H GTC TTC CAG CTG GAG AAG GGT GAC CGA CTC AGC GCT GAG ATC PST
R CAG CCT GAG TAC CTG GAC CTT GCC GAG TCC GGG CAG GTC TAC TTT
H CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
R GGG ATC ATT GCC CTG
H GGG ATC ATT GCC CTG
Note: the symbol "..." means that this portion in the base sequence o:f the DNA coding for rabbit TNF is~rull and, therefore, two codons adjacent to this symbol at its both sides are directly connected.

s~e~ la (Synthesis of oligodeos:_rn::cleotices) To a stainless steel 500 ~1 reactio;~ vessel with stainless steel filters at each ezd is added 20 mg of a poly-styrene resin to wi-:ich a nucleoside (2.0 yM) is connected via a succinate linkage. The resin is t=eaten with zinc bromide (1M) in dichloromethane isopropanol(85:15) to remove the dimetho::ytrityl (DP~1T) protecting group, washed with dimethyl-formamide, pyridine, and acetonitrile, and dried with a stream of nitrogen. To the dried resin is added a solution of DMT-nucleotide (20 ~uM) and mesityle~esul~onylnitrotriazole ..
(60 ~L~i) in 200 ~1 p~~ridine. The couplinc react: on is allo«ed to procee~ at 45'C zor 20 minutes. This cycle of deprotection and coupling is repeated for successive nucleotides until the a°-sired olicodeoxynucleotide is assz:~~bled on the rasin. The resin is then t=eared to re:r:oce the olicodeoxvnucieotide there'rom and purified as describe3 by Ito, Ike, Ik~,:ta, and Itakura (Nuc. Ac. Res. 10:1 755 (1982) ) .
Thus, the following olicodeoYya~uc:eotices are o;,taine3.
1) 5'-AATTCATGTCATC:TTCTCGAACCCCGAGTGACAA-3' 2) 3'-GTACAGTAG=~ GAGCTTGGCGCTCACTGTTCGG-5' 3) 5'-GCCTGTAGCCCATGTTGTAGCAAACCCTCAAGC-3"
4) 3'-ACATCGGGTC~.ACATCGTTTGGGAGTTCGACT-5°
Step 15 (Construction of hi13mp9-HGE containing the human minigene for TNF) Plasmid pHGE (10 Yg) is digested with EcoRI(20 units).
After electrophoresis on a 1 % low-melting aaarose gel, the 2.9 kb fragment is eluted. This ~ragment is inse=ted into EcoRI fragment from the replicative form of M13mp9 phage. The M13mp9 phage is selected because it is especial-ly suited for receiving sections of DNA. The product trans-fects to E. coli JrI103 (BRL (Bethesda Research Laboratories, Inc., U.S.A.) User Manual/M13mp7 Cloning/'Dideoxy' seQUencing, 19801. The product is designated Mi3mp9-HGE.
Step 16 (Deletion of Intron 3, using M13mp9-hGE
single strand DNA and Deletes E3-4) The s i ncle strand DNA of l~il3:mp9-iiGE is prepared by the method of BRI, User r~.::,:~1/r13 ~7 clcuinc,/'Diceoxy' sue'~.cina, 1980.
Olicodeox~nucieotide 4) 3' -~CATCGGG'ACA.~.CATCGTTTGGG?GTTCGACT-5' prsca=ed in St°-p 14 is used as a deletes for t:.e i:.t=on 3.
The deletes for the intron 3 is designated "E3-4".
The deletes E3-4 has a base sequence which is complementary to the base sequence of the bases be~or2 (Exon 3) and after (Exon 4) the intron 3 which is to be deleted. Deletion of the intson 3 is effected, in accordance with the teaching of t~~llace et al, Science 209:1396 (1980), as follows.
E3-4(164 ng, 15 pmole) is phosphorylated using T4 kinase and ATP (3mM) and added to the template M13mp9-HGE (1.65 Yg, 0.5 pmole). The reaction mixture is heated at 65oC ~~~~.-m,~.-R~t~e~-, cooled to room temperature for 5 minutes , and finally cooled in ice water. To dATP, dCTP, dGTP, dTTP and ATP (0.4 mM), is added Klenow fragment (5 units), T4 ligase (10 units) in Hin buffer [Wallace et al, I~uc. Ac. P.es. 2;
3647 (1981) ] , 10 mM Tris HCl (pH 7.2) , 2 mt~l MgCl2 and 1mM
8-mercaptoethanol. The reaction mixture (final volume 50 ~1) is incubated for 30 minutes at 4oC and then for 30 minutes at room temperature. The DNA from the oligonucleo~tide-grimed reaction is used to transfect E. coli JM103 in accordance with the procedure of BRL User Manual/M13mp7 cloning/'Didevxy' sequencing, 1980. Plaques obtained in this way are picked to YT
plates [J. H. Miller, p. 433, Experiments in Molecular Genetics, Cold~Spr_ng Harbor Laboratory (1Q72)]. '"he colonies obtained are hybridized at 55oC for 2 hours with 32P-labelled E3-4. For this step, the deleter is used as a probe 1.o identify sequences of DNA having the corresponding complementary base sequence of ter the intron has been deleted. Phage are isolated from those colonies which hybridize with the deleter.
The resultant phage are plated and pl~:ques are picked to YT
plates. The clones are allowed to hybridi~:e at 55oC for 2 hours with 32P-labelled E3-4. Positive clones are obtained and the phage DNA is sequenced to select those phage in which intron 3 is completely deleted. One such phage is designated mpg-AGE
3-1.
_ 78 -~ 341 348 Step 17 ( Construction of pHT:3F-laCUVS-2).
The replicatil~e form of mD9-HGE~3-1 is digested with EcoRT. The EcoRI fragment is ~.solat:ed and cloned to EcoRI-cleaved pBR327 to yield the plasmid pHGE C 3-1.
Construction of further p.lasmid is carried out using plasmid pHGE d 3-1 in order to obtain such plasmid ~3-1 as will directly e:cpress TNF in E. coli using lac W5 as a promoter.
The procedures are illustratively shown in Fig. 7. First, 10 ug of plasmid pHGE ~- 3-1 is digested with 10 units of AvaII and EcoRI (manufactured and sold by Bethesda Research Laboratories, Ins., U.S.A.) at 37°C for two hours and electro-phoresed on a 4 o by weight polyacrylamide gel to isolate fraGments. About 1. uQ of fragment is isolates from the _ ~ ., _ gel by electroelution. In the same manner as in Step 14,' two olicodeoxvnucleotides shown in Fig. 7, namely 5'-AATTC -TGTC."-~TCTTC':.'CGr~CC-3' anC 5'-TCGG:~GTTCG~G.-~~AGATGnC TG-3' art Synthesized. The.~., e~cn 5' end Of the two 011gOCe0yy-nucleotides (about 100 pmole) is phasphorylated using T4 polynucleotide kinase in accordance with the method described In Llteratune (3), page 1~~. A=ter COmDIe'tl0n Of the re3CtlOn, the reaction miWure is extracted with phenol and then with chloroform. Then '-~.i:e so-obt~ine;i s,~~nthetic oligom~rs are ~'~i.'~e3 wiy~h 0.5 ~g oL tine previously c~~ainea AvaI-EcoRI =racment from plas.-nid pHGE J 3-1 and ethanol precipitated. These fragments are ligated at 4°C overnight using 10 units of T4 ligase in accordance with the procedure described in.Literature (1), - 79 _ page 37. After completion of the reaction, the mixture is ethanol precipitated, followed by electrophoresis effected on a 4 % by weight poiyacrylamide gel to recover.fragment by electroelution.
In accordance with tre procedure described in F. Fuller, "Gene", 1S, pp. 42-54 (1982), plasmid pOP95-15 is prepared.
One ug of pOP 95-15 is digested with EcoRI and extracted with phenol and then with chloroform, followed by ethanol precipitation to obtain a vector. Using T4 DNA lipase, 0.5 ug of the obtained vector is ligated with the above-obtained fragment. In accordance with the procedure described in Literature (4) , page 20, E. coli Jt4101 (ATCC 33876) is transformed using the above-obtained vector and cultivated on an agar -medium containing 1 mM of IPTG and 0.004 w/v o x-gal to obtain about 100 white colonies.
Plasmids DNA is prepared from these transf ormants and digested with EcoP,I to identify those plasmids containing the intended EcoRI fragment. In order to examine the direction of insertion, those plasmids are digested with PvuII and PstI and electrophoresed on a 1.5 % by weight agarose gel to select plasmids yielding fragments of about 1280 base pairs and about 2500 base pairs indicating that the direction of transcription of the lac W5 promoter is in agreement with those of the oligodeoxynucleotides coding f or TNF.
_ 80 -Base sequence analysis shows that these 2 plasmids have the same sequence and that the lac W5 promotoer, the synthesized oligodeoxynucleotide and cDNA are properly combined with each .. lasini c~.
other. The obtained ~~~:.s is designated pHTNF-lacWS-2.
E, coli containing pHT~:F-lacWS-2 is cultured in a conventional nutrient medium. Bioassay of the product for TNF activity indicates almost the same activity whihe is obtained with a plasr~id pTNF-lacUVS-1 containing the rabbit TNF gene under control of the lac promoter.o Exam'_ole ~2 Using the plasmid pHGE and oligodeoxynucleotides 1 to 4) obtained by the procedure described in Steps 1 to la of Example 1 pHTNF-lacWS-1 is prepared in accordance with the procedure illustrated in Fig.W

The microorganisms and the novel plasmids were placed on deposit in the American Type Culture Collection in Rockville, Maryland. U.S.A. on April 6. 1984 by despositing samples of the microorganisms containing the plasmids. The microorganism ~.
coli k-12 strain JM83 (pRGE) Was given the A'rCC accession number 39655. The microorganism ~..~i k-12 strain JM83 (pHGE) was given the ATCC accession number 39656.
The invention being thus described, it will be obcTious that the same may be varied in many ways. Such variations are not to be regarded as a departure from the spirit and scope of the invention, and all such modifications as would be obvious to one skilled in the art are intended to be included within the scope of the following claims.

Claims (30)

1. A deoxyribonucleic acid capable of coding for a human polypeptide having tumor necrosis factor activity, comprising at least one base sequence selected from the group consisting of a base sequence represented by the following formula (II) and a complementary base sequence to said base sequence:
TCA TCT TCT CGA ACC CCG AGT GAC AAG CCT GTA GCC CAT GTT GTA
GCA AAC CCT CAA GCT GAG GGG CAG CTC CAG TGG CTG AAC CGC CGG
GCC AAT GCC CTC CTG GCC AAT GGC GTG GAG CTG AGA GAT AAC CAG
CTG GTG GTG CCA TCA GAG GGC CTG TAC CTC ATC TAC TCC CAG GTC
CTC TTC AAG GGC CAA GGC TGC CCC TCC ACC CAT GTG CTC CTC ACC
CAC ACC ATC AGC CGC ATC GCC GTC TCC TAC CAG ACC AAG GTC AAC
CTC CTC TCT GCC ATC AAG AGC CCC TGC CAG AGG GAG ACC CCA GAG
GGG GCT GAG GCC AAG CCC TGG TAT GAG CCC ATC TAT CTG GGA GGG
GTC TTC CAG CTG GAG AAG GGT GAC CGA CTC AGC GCT GAG ATC AAT
CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
GGG ATC ATT GCC CTG
wherein A stands for a deoxyadenylic acid residue, G a deoxyguanylic acid residue, C a deoxy-cytidylic acid residue and T a thymidylic acid residue and wherein the left end and right end of the formula (II) represent 5'-hydroxyl group side and 3'-hydroxyl group side, respectively, or com-prising a base sequence which is obtained by sub-stituting at least one base of said base sequence in accordance with degeneracy of genetic code.

2. A replicable recombinant DNA which comprises a deoxyribo-nucleic acid according to claim 1, and a replicable expression vehicle.

3. An E. coli transformed with a replicable recombinant DNA
according to claim 2.

4. A method for producing a human physiologically active polypeptide having tumor necrosis factor activity which comprises:
(a) ligating a DNA which codes for a human physiologi-cally active polypeptide having tumor necrosis factor activity to a replicable expression vector which is capable of expressing said DNA to obtain a replicable recombinant DNA
comprising said DNA and said replicable expression vehicle, said human physiologically active polypeptide being a polypeptide having an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu G1u Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a Tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue, or being a homologous variant of said polypeptide having TNF
activity;
(b) transforming an E. coli with said replicable recom-binant DNA to form transformants;
(c) selecting said transformants from-parent cells of the E. coli;
(d) incubating said transformants, causing said trans-formants to express said DNA and produce a human physiologi-cally active polypeptide; and (e) isolating said human physiologically active polypep-tide from the incubated transformants.

5. A method according to claim 4, wherein said DNA is a DNA

comprising at least one base sequence selected from the group consisting of a base sequence represented by the following formula (II) and a complementary base sequence to said base sequence:
TCA TCT TCT CGA ACC CCG AGT GAC AAG CCT GTA GCC CAT GTT GTA
GCA AAC CCT CAA GCT GAG GGG CAG CTC CAG TGG CTG AAC CGC CGG
GCC AAT GCC CTC CTG GCC AAT GGC GTG GAG CTG AGA GAT AAC CAG
CTG GTG GTG CCA TCA GAG GGC CTG TAC CTC ATC TAC TCC CAG GTC
CTC TTC AAG GGC CAA GGC TGC CCC TCC ACC CAT GTG CTC CTC ACC
CAC ACC ATC AGC CGC ATC GCC GTC TCC TAC CAG ACC AAG GTC AAC
CTC CTC TCT GCC ATC AAG AGC CCC TGC CAG AGG GAG ACC CCA GAG
GGG GCT GAG GCC AAG CCC TGG TAT GAG CCC ATC TAT CTG GGA GGG
GTC TTC CAG CTG GAG AAG GGT GAC CGA CTC AGC GCT GAG ATC AAT
CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
GGG ATC ATT GCC CTG
wherein A stands for a deoxyadenylic acid residue, G a deoxyguanylic acid residue, C a deoxycytidylic acid residue and T thymidylic acid residue and wherein the left end and right end of theformula (II) represent 5'-hydroxyl group side and 3'-hydroxyl group side, respectively.

6. A method for producing a human physiologically active polypeptide having tumor necrosis factor activity which comprises:
(a) ligating a DNA to a replicable expression vehicle which is capable of expressing said DNA to obtain a replicable recombinant DNA comprising said DNA and said replicable ex-pression vector, said DNA comprising a first DNA coding for a human physi-ologically active polypeptide having tumor necrosis factor activity and a second DNA coding for a peptide and ligated to the 5'-end of said first DNA, said human physiologically active polypeptide being a polypeptide having an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val ALA Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Sex Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a Tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue, or being a homologous variant of said polypeptide having TNF
activity, and said second DNA being one selected from the group con-sisting of:
(i) a DNA coding for a signal peptide which is derived from a microorganism and has an amino acid sequence composed of from 15 to 40 amino acids;
(ii) a DNA coding for signal peptide which is derived from a higher animal and has an amino acid sequence composed of from 15 to 40 amino acids; and (iii) a DNA coding for a peptide having an amino acid sequence corresponding to part of an intermediate form of said human physiologically active polypeptide, said part being the remainder of said intermediate form from which said human physiologically active polypeptide is removed;
(b) transforming an E. coli with said replicable recom-binant DNA to form transformants;
(c) selecting said transformants from parent cells of the E. coli;

(d) incubating said transformants, causing said trans-formants to express said DNA and produce a polypeptide product having an amino acid sequence portion corresponding to a human physiologically active polypeptide, the C-terminal amino acid of said amino acid sequence portion being in coincidence with the C-terminal amino acid of said polypeptide product; and (e) isolating said polypeptide product from the incubat-ed transformants.

7. A method according to claim 6, wherein said polypeptide product has an amino acid sequence portion corresponding to said peptide derived from said second DNA and bonded to the N-terminus of said amino acid sequence portion corresponding to said human physiologically active polypeptide and which further comprises chemically or enzymatically cleaving said polypeptide product to obtain said human physiologically active polypeptide and said peptide derived from said second DNA and separating said physiologically active polypeptide from said peptide derived from said second DNA.

8. A method according to claim 6, wherein said first DNA is a DNA comprising at least one base sequence selected from the group consisting of a base sequence represented by the follow-ing formula (II) and a complementary base sequence to said base sequence:

TCA TCT TCT CGA ACC CCG AGT GAC AAG CCT GTA GCC CAT GTT GTA
GCA AAC CCT CAA GCT GAG GGG CAG CTC CAG TGG CTG AAC CGC CGG
GCC AAT GCC CTC CTG GCC AAT GGC GTG GAG CTG AGA GAT AAC CAG
CTG GTG GTG CCA TCA GAG GGC CTG TAC CTC ATC TAC TCC CAG GTC
CTC TTC AAG GGC CAA GGC TGC CCC TCC ACC CAT GTG CTC CTC ACC
CAC ACC ATC AGC CGC ATC GCC GTC TCC TAC CAG ACC AAG GTC AAC
CTC CTC TCT GCC ATC AAG AGC CCC TGC CAG AGG GAG ACC CCA GAG
GGG GCT GAG GCC AAG CCC TGG TAT GAG CCC ATC TAT CTG GGA GGG
GTC TTC CAG CTG GAG AAG GGT GAC CGA CTC AGC GCT GAG ATC AAT
CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
GGG ATC ATT GCC CTG
wherein A stands for a deoxyadenylic acid residue, G a deoxyguanylic acid residue, C a deoxycytidylic acid residue and T thymidylic acid residue and wherein the left end and right end of the formula (II) represent 5'-hydroxyl group side and 3'-hydroxyl group side, respectively.

9. A method for producing a human physiologically active polypeptide having tumor necrosis factor activity which comprises:
(a) ligating a first DNA which codes for a human physio-logically active polypeptide having tumor necrosis factor activity to a replicable expression vehicle which contains a DNA coding for a predetermined peptide and is capable of expressing said first DNA and said DNA coding for said predetermined peptide, thereby to obtain a replicable recombinant DNA, said human physiologically active polypeptide being a polypeptide having an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a Tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue, or being a homologous variant of said polypeptide having TNF
activity;

(b) transforming an E. coli with said replicable recom-binant DNA to form transformants;
(c) selecting said transformants from parent cells of the E. coli;
(d) incubating said transformants, causing said trans-formants to express said first DNA and said DNA coding for said predetermined peptide and produce a fused peptide com-prising said human physiologically active polypeptide and said predetermined peptide;
(e) isolating said fused peptide from the incubated transformants; and (f) chemically or enzymatically cleaving said fused peptide to obtain said human physiologically active polypep-tide and said predetermined peptide, and separating said human physiologically active polypeptide from said predetermined peptide.

10. A method according to claim 9, wherein said first DNA is a DNA comprising at least one base sequence selected from the group consisting of a base sequence represented by the following formula (II) and a comple-mentary base sequence to said base sequence:
TCA TCT CGA ACC CCG AGT GAC AAG CCT GTA GCC CAT GTT GTA GCA
GCA AAC CCT CAA GCT GAG GGG CAG CTC CAG TGG CTG AAC CGC CGG
GCC AAT GCC CTC CTG GCC AAT GGC GTG GAG CTG AGA GAT AAC CAG

CTG GTG GTG CCA TCA GAG GGC CTG TAC CTC ATC TAC TCC CAG GTC
CTC TTC AAG GGC CAA GGC TGC CCC TCC ACC CAT GTG CTC CTC ACC
CAC ACC ATC AGC CGC ATC GCC GTC TCC TAC CAG ACC AAG GTC AAC
CTC CTC TCT GCC ATC AAG AGC CCC TGC CAG AGG GAG ACC CCA GAG
GGG GCT GAG GCC AAG CCC TGG TAT GAG CCC ATC TAT CTG GGA GGG
GTC TTC CAG CTG GAG AAG GGT GAC CGA CTC AGC GCT GAG ATC AAT
CGG CCC GAC TAT CTC GAC TTT GCC GAG TCT GGG CAG GTC TAC TTT
GGG ATC ATT GCC CTG
wherein A stands for a deoxyadenylic acid residue, G a deoxyguanylic acid residue, C a deoxycytidylic acid residue and T thymidylic acid residue and wherein the left end and right end of the formula (II) represent 5'-hydroxyl group side and 3'-hydroxyl group side, respectively.

11. A method according to claim 4, 6 or 9,.wherein said transformants are so incubated that said transformants are multiplied and subsequently said transformants are caused to express said DNA and produce a human physiologically active polypeptide.

12. A method for producing a human physiologically active polypeptide having tumor necrosis factor activity which comprises:
(a) providing a transformant E. coli, having been trans-formed with a replicable recombinant DNA comprising a DNA
coding for a human physiologically active polypeptide having TNF activity and a replicable expression vehicle, said human physiologically active polypeptide being a polypeptide having an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a Tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue, or being a homologous variant of said polypeptide having TNF

activity;
(b) incubating said transformant, causing said trans-formant to express said DNA and produce a human physiologically active polypeptide; and (c) isolating said human physiologically active polypeptide from the incubated transformant.

13. A method according to claim 12, wherein said transformant is so incubated that said transformant is multiplied and subsequently said transformant is caused to express said DNA and produce a human physiologically active polypeptide.

14. A substantially pure physiologically active non-glycosylated polypeptide having human TNF activity prepared by the method claimed in claim 4.

15. A substantially pure physiologically active non-glycosylated polypeptide having human TNF activity prepared by the method claimed in claim 6, 7 or 9.

16. A substantially pure physiologically active non-glycosylated polypeptide having human TNF activity prepared by the method claimed in claim 12 or 13.

17. A human physiologically active non-glycosylated polypeptide comprising an amino acid sequence represented by the following formula (I):
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Tyr Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu wherein Gln stands for a glutamine residue, Asp an aspartic acid residue, Pro a proline residue, Tyr a Tyrosine residue, Val a valine residue, Lys a lysine residue, Glu a glutamic acid residue, Ala an alanine residue, Asn an asparagine residue, Leu a leucine residue, Phe a phenylalanine residue, Gly a glycine residue, His a histidine residue, Ser a serine residue, Thr a threonine residue, Ile an isoleucine residue, Trp a tryptophan residue, Arg an arginine residue, Met a methionine residue, and Cys a cysteine residue, or a homologous variant of said polypeptide having TNF activi-ty, when produced by the method claimed in claims 5 or 8.

18. A plasmid or bacteriophage transfer vector comprising a base sequence coding for a polypeptide defined in claim 14.

19. An E. coli transformed by the transfer vector of claim 18, and mutants or variants thereof.

20. The microorganism E, coli k-12 strain JM83 (pRGE).

21. The microorganism E. coli k-12 strain JM83 (pHGE).

22. The plasmid pRGE.

23. The plasmid PHGE.

24. A pharmaceutical composition comprising an effective anti-tumor or anti-viral amount of polypeptide as defined in claim 14 and a pharmaceutically acceptable carrier, diluent or excipient.

25. A polypeptide defined by the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val G1u Leu Arg Asp Asn Gln Leu Va1 Val Pro Ser Glu Gly Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu.

26. Recombinant DNA encoding a polypeptide defined by the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val A1a Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg A1a Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe G1n Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu.

27. Expression vector comprising a DNA sequence encoding a polypeptide defined by the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu.

28. Transformed host cell comprising an expression vector comprising a DNA encoding a polypeptide defined by the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val A1a His Val Val Ala Asn Pro Gln Ala Glu G1y Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu G1y Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu.

29. Process for the preparation of a human tumor necrosis factor comprising transforming a host cell with an expression vector containing a DNA sequence encoding the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly Gln Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu, culturing the transformed host cell and collecting the tumor necrosis factor expressed by said transformed host cell.

30. Use of a polypeptide defined by the amino acid sequence:
Ser Ser Ser Arg Thr Pro Ser Asp Lys Pro Val Ala His Val Val Ala Asn Pro Gln Ala Glu Gly Gln Leu Gln Trp Leu Asn Arg Arg Ala Asn Ala Leu Leu Ala Asn Gly Val Glu Leu Arg Asp Asn Gln Leu Val Val Pro Ser Glu Gly Leu Try Leu Ile Tyr Ser Gln Val Leu Phe Lys Gly G1n Gly Cys Pro Ser Thr His Val Leu Leu Thr His Thr Ile Ser Arg Ile Ala Val Ser Tyr Gln Thr Lys Val Asn Leu Leu Ser Ala Ile Lys Ser Pro Cys Gln Arg Glu Thr Pro Glu Gly Ala Glu Ala Lys Pro Trp Tyr Glu Pro Ile Tyr Leu Gly Gly Val Phe Gln Leu Glu Lys Gly Asp Arg Leu Ser Ala Glu Ile Asn Arg Pro Asp Tyr Leu Asp Phe Ala Glu Ser Gly Gln Val Tyr Phe Gly Ile Ile Ala Leu as an anti-tumor agent.