WO1996013586A2

WO1996013586A2 - Histiocyte-secreted factor (hsf)

Info

Publication number: WO1996013586A2
Application number: PCT/JP1995/002200
Authority: WO
Inventors: Nobuko Satomi
Original assignee: Nobuko Satomi
Priority date: 1994-10-26
Filing date: 1995-10-26
Publication date: 1996-05-09
Also published as: WO1996013586A3

Abstract

The present invention provides isolated nucleotide sequences encoding histiocyte-secreted factor (HSF) or an HSF analog and the complements of these sequences. Also provided are isolated nucleotide sequences comprising a portion of an HSF genomic sequence, which can consist of coding regions, non-coding regions, or both. The invention includes nucleotide sequences suitable for use as hybridization probes and amplification primers and for insertion into vectors for propagation and expression. The invention also provides isolated HSF polypeptides and analogs, along with anti-HSF antibodies. These polypeptides and antibodies are useful in cancer chemotherapy and as tools in the study of tumor regression.

Description

HISTIOCYTE-SECRETED FACTOR (HSF), A NOVEL CYTOKINE

Nobuko Satomi

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates to a novel cytokine, termed "histiocyte-secreted factor" (HSF), which has anti-tumor activity. More particularly, the invention relates to nucleotide sequences encoding HSF, HSF polypeptides, and anti-HSF antibodies.

Description of the Related Art

Cytokines are a family of proteins with a wide variety of biological actions, many of which are related to the body's response to invasive stimuli. A number of cytokines having anti-tumor activity have been studied as potential cancer chemotherapeutics; these include interferon, tumor necrosis factor-α(TNF), lymphotoxin, and various interleukins. TNF, for example, exhibited excellent anti-tumor activity in animal models. Haranaka et al., 34 Int. J. Cancer 263-267 (1984).

Unfortunately, however, recombinant TNF was highly toxic in animal studies (Satomi et al., 7 J. Biol. Response Mod. 54-64 (1988)) and in human clinical trials. Systemic

administration of recombinant TNF induced shock-related symptoms, which in some instances resulted in death.

Furthermore, administration of the maximum-tolerated dose of recombinant TNF to humans was found to have no appreciable therapeutic effect, probably because this dose was only 2% of the dose required for tumor regression in mice. Despite these difficulties, efforts to develop TNF as a cancer chemotherapeutic continue, focusing on the administration of TNF in combination with agents that protect against TNF-induced toxicity. TNF's toxicity appears to be mediated, at least in part, by the generation of oxygen radicals via the arachidonic acid cascade. Therefore, the co-administration of TNF with oxygen radical scavengers or drugs that interfere with the production of arachidonic acid or its metabolites has been investigated.

Ideally, however, a cytokine useful in cancer treatment would exhibit good anti-tumor activity without the severe toxicity observed with TNF. SUMMARY OF THE INVENTION

The present invention relates to a novel cytokine, termed histiocyte-secreted factor (herinafter "HSF"), having in vivo anti-tumor activity. As in vitro

cytotoxicity studies have demonstrated that HSF is far less toxic than TNF, HSF is useful as a cancer

chemotherapeutic agent, in addition to having utility as a tool for studying the basis of tumor regression. The invention provides isolated HSF nucleotide sequence, HSF polypeptides and analogs, and anti-HSF antibodies.

In a first embodiment, a nucleotide sequence of the present invention encodes, or is complementary to a sequence encoding, HSF or an analog thereof. The

nucleotide sequence can be a genomic DNA sequence, a cDNA sequence, or an mRNA sequence.

In particular, the nucleotide sequence can encode human HSF. In a variation of this embodiment, the

nucleotide sequence encodes a polypeptide comprising an amino acid sequence as set forth in Table 1. In a second variation of this embodiment, the encoded polypeptide polypeptide has the amino acid sequence set forth in Table 1.

As HSF analog of this invention can comprise a mutation of a naturally occurring histiocyte-secreted factor consisting of a conservative amino acid

substitution.

Also provided is an anti-HSF antibody, which binds the HSF polypeptide and/or HSF analog of this invention. Such antibodies are useful for purifying or detecting an HSF polypeptide or analog.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1A shows the elution pattern observed when serum from rabbits inoculated with a Propionibacterium acnes vaccine was subjected to Q Sepharose Fast Flow column chromatography, eluting with a linear NaCl gradient of from 0.15 M to 1.00 M in 40 mM Tris buffer (pH 7.8; fine line in FIG. 1A). The amount of protein in each fraction was monitored by absorbance (OD₂₆₀; thick line in FIG. 1A).

FIG. 1B shows the results of a study in which the in vivo anti-tumor activity of Q Sepharose fractions was evaluated using transplanted murine Meth A sarcomas.

FIG. 2 shows the elution pattern ooserved when peptides generated from purified rabbit HSF by digestion with acromobacter protease were subjected to reverse phase high performance liquid chromatography (HPLC), eluting with a linear acetonitrile gradient of from 0% to 72% in the presence of 0.1% trifluoroacetate.

FIG. 3A shows an immunostaining of partially purified human and rabbit HSF using anti-rabbit HSF peptide antibodies (anti-13-Ab and anti-7-Ab), compared with a Coomassie blue-stained SDS-PAGE gel of partially purified human and rabbit HSF electrophoresed with molecular weight markers (CBB).

FIG. 3B shows the results of a densitometer scan of the immunostaining of partially purified human and rabbit HSF using the anti-13mer-Ab shown in FIG. 3A.

FIG. 4 shows the results of PEPPLOT analysis, which plots measures of protein secondary structure and hydrophobicity in parallel panels of the same plot.

FIG. 5 shows the results of Helical wheel

projections of alpha helices in HSF. Hydrophobic amino acid residues are boxed.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is based on the discovery of a novel cytokine, histiocyte-secreted factor (hereinafter

"HSF"), that exhibits in vivo anti-tumor activity without the cytotoxicity that has frustrated attempts to develop TNF as a cancer chemotherapeutic. This invention provides isolated nucleotide sequences encoding HSF or an HSF analog and the complements of these sequences. Also provided are isolated nucleotide sequences comprising a portion of an HSF gene sequence, which can consist of coding regions, non-coding regions, or both. The

invention includes nucleotide sequences suitable for use as hybridization probes and amplification primers and for insertion into vectors for propagation and expression.

The invention also provides isolated HSF polypeptides and analogs, along with anti-HSF antibodies. Nucleotide Sequences

The isolated nucleotide sequences of the present invention are generally less than 100 kilobases (kb), conveniently less than 50 kb, more conveniently less than 10 kb, and most conveniently less than 2 kb. The consists essentially of the amino acid sequence set forth in Table 1. In a third variation, the polypeptide has the amino acid sequence set forth in Table 1.

The nucleotide sequence of this embodiment can be a naturally occurring HSF nucleotide sequence or a modified sequence. In a variation of this embodiment, the

nucleotide sequence is that set forth in Table 1.

A nucleotide sequence encoding an HSF analog of this invention can comprise a mutation consisting of a

conservative amino acid substitution.

In a second embodiment, a nucleotide sequence of the present invention comprises a portion of an HSF genomic sequence. Such sequences are useful, for example, as hybridization probes and/or amplification primers.

The present invention also provides a vector

comprising a nucleotide sequence encoding HSF or an analog thereof and control sequences that are capable of effecting expression of this nucleotide sequence in a suitable host cell . Also provided is a vector comprising a nucleotide sequence including a portion of a

histiocyte-secreted factor genomic sequence and a

replication sequence that is capable of effecting

replication of the vector in a suitable host cell. A host cell according to the invention is transformed with one of these vectors.

The present invention additionally provides an isolated polypeptide selected from the group consisting of an HSF, polypeptide or an analog thereof. In

particular, the polypeptide can be human HSF. In a

variation of this embodiment, the polypeptide comprises an amino acid sequence as set forth in Table 1. In a second variation, the polypeptide consists essentially of this amino acid sequence, and in a third variation, the nucleotide sequences can comprise HSF genomic DNA, cDNA, or mRNA sequences. Genomic DNA includes non-transcribed regions flanking the HSF gene as well as transcribed regions, including 5' and 3' non-coding regions, introns, and HSF coding regions. cDNA and mRNA sequences include the transcribed regions found in mature HSF mRNA.

Nucleotide Sequences Encoding HSF and HSF Analogs

In a first embodiment, a nucleotide sequence of the present invention encodes HSF or an HSF analog. This nucleotide sequence encompasses sequences comprising a portion of the HSF or HSF analog coding region as well as sequences comprising the entire HSF or HSF analog coding region. Generally, the nucleotide sequence comprises sequences encoding the N-terminal 25%, conveniently 50%, more conveniently 75%, of the HSF polypeptide or analog.

In one variation of this embodiment, the encoded HSF is a naturally occurring HSF from any species, generally a vertebrate, conveniently a mammal. Nucleotide

sequences of this invention encoding human HSF, for example, include a nucleotide sequence encoding the human HSF amino acid sequence set forth in Table 1, as well as allelic variants thereof.

*Isolated from lambda gtll library prepared from human U-937 monocytic cells stimulated with phorbolmyristate acetate

(The abbreviations for the nucleotides are as follows: G, guanine; A, adenine; C, cytosine; T, thymine.

The abbreviations for the amino acids are as follows: Gly, glycine; Ala, alanine; Val, valine; Leu, leucine; Ile, isoleucine; Ser, serine; Thr, threonine; Asp, aspartic acid; Glu, glutamic acid; Asn, asparagine; Gin, glutamine; Lys, lysine; Arg, arginine; Cys, cysteine; Met, methionine; Phe, phenylalanine; Tyr, tyrosine; Trp, tryptophan; His, histidine; Pro, proline.)

The nucleotide sequence can be a naturally occurring HSF nucleotide sequence, such as, for example, the human HSF nucleotide sequence set forth in Table 1.

Alternatively, the nucleotide sequence can be a

nucleotide sequence that differs from a naturally occurring HSF nucleotide sequence, but still encodes the same HSF polypeptide. Such nucleotide sequences can comprise a nucleotide insertion, deletion, or

substitution in a non-coding region. Techniques for introducing such mutations into nucleotide sequences are well-known and include, for example, site-directed or cassette mutagenesis, as well as direct synthesis of mutated sequences. In addition, because of the well-established redundancy of the genetic code, such nucleotide sequences can also comprise a nucleotide substitution in the coding region. More specifically, because several different codons encode the same amino acid, a codon substitution need not produce an amino acid substitution. Examples of such "silent mutations" within the scope of the present invention include the substitution of a codon "preferred" by a selected host to facilitate translation of mRNA in that host or the creation or destruction of a restriction endonuclease site to facilitate construction of a desired vector.

In another variation of this embodiment, a

nucleotide sequence of the present invention encodes an HSF analog. An HSF analog is a modified HSF polypeptide that is at least 60%, generally 80%, conveniently 90%, homologous to a naturally occurring HSF. Exemplary modifications include conservative amino acid

substitutions anywhere in the HSF amino acid sequence and/or insertions and/or deletions in non-conserved regions of HSF.

Non-conserved regions are determined conventionally by comparing the amino acid sequences of HSF from different species.

The nucleotide sequence of the present invention can also be a sequence complementary to any of the nucleotide sequences described above. A complementary nucleotide sequence is capable of forming Watson-Crick bonds with its complement, in which adenine pairs with thymine or uracil and guanine pairs with cytosine. In addition, the nucleotide sequence can be a sequence that hybridizes to any of the above-described nucleotide sequences with a melting temperature (T_m) above 55°C, generally in the range of about 60°C to about 80°C. Conveniently, the T_m is from about 65°C to about 70°C to facilitate hybridization under stringent conditions.

HSF Genomic Sequences

In a second embodiment, a nucleotide sequence of the present invention comprises a portion of an HSF genomic sequence. Such sequences find use as hybridization probes for detecting HSF genes or closely-related genes in human or other species and, if the genomic sequence includes a transcribed region, for detecting HSF RNA. In addition, an isolated nucleotide sequence of this

embodiment is useful as an amplification primer in

conventional amplification protocols such as polymerase chain reaction (PCR). Accordingly, the present invention provides a hybridization probe that is capable of hybridizing with a transcribed region and/or untranscribed region of an HSF gene. The latter type of probe is useful in isolating HSF genomic sequences, such as, for example, HSF promoter sequences. A probes that hybridizes with an HSF coding region is also provided. Such probe is useful for detecting mRNA as well as DNA sequences.

A probe of the present invention can range in length from at least about 20 nucleotides to not more than about 50 kb, and usually less than 30 kb. The probe includes a number of nucleotides that is sufficient, under the hybridization conditions used, to hybridize with the sequence to be detected and to be substantially free from hybridization with other sequences. Typically, a probe of the present invention has at least about 50,

conveniently about 100, more conveniently about 500 nucleotides complementary to the sequence to be detected. In one embodiment, the probe includes at least about 20 consecutive nucleotides complementary to the sequence to be detected.

The present invention also provides an amplification primer which is typically used as a member of a primer pair. A primer pair includes a 5' upstream primer that hybridizes with the 5' end of the nucleotide sequence to be amplified and a 3' downstream primer that hybridizes with the complement of the 3' end of the sequence to be amplified.

In general, each primer includes a number of

nucleotides that is sufficient, under the hybridization conditions used, to hybridize with a sequence flanking (hereinafter "target sequence") the sequence to be

amplified and to be substantially free from hybridization with other sequences. The specificity of the primer increases with the number of nucleotides that hybridize with the target sequence. Therefore, longer primers are desirable. A primer of the present invention generally includes at least about 15 nucleotides, conveniently at least about 20 nucleotides. The primer need not exceed about 30 nucleotides, and conveniently does not exceed about 25 nucleotides. In one variation of this

embodiment, the primer includes between about 20 and about 25 nucleotides.

A primer of the present invention generally has a nucleotide sequence that is complementary the target sequence. However, a primer of 20-25 nucleotides can have two non-complementary nucleotides, and longer primers can have more. Any non-complementary nucleotides are conveniently not located at the 3' end of the primer. In one variation of this embodiment, the 3' end of the primer has at least two, conveniently three or more, nucleotides that are complementary to the target

sequence. Additionally, any non-complementary

nucleotides are conveniently not adjacent in the primer sequence. In one variation of this embodiment, any non- complementary nucleotides are separated by at least three, more preferably at least five, nucleotides. The primers should have a melting temperature (T_m) in the range of about 55°C to 75°C. Generally, the T_m is from about 60°C to about 65°C to facilitate amplification under stringent hybridization conditions.

Production of Nucleotide Sequences

The isolated nucleotide sequences of the present invention can be produced by purifying genomic DNA comprising HSF gene sequences, by purifying HSF mRNA, or by using standard techniques for preparing cDNA. The equivalent of genomic or cDNA sequences can also be produced by amplifying genomic DNA or mRNA, respectively. Alternatively, the isolated nucleotide sequence of the present invention can be directly synthesized. Methods for isolating cDNA and genomic sequences, amplifying mRNA and DNA, and synthesizing DNA are well known and

constitute no part of this invention.

The isolation of a genomic DNA sequence comprising HSF coding regions is described in Examples 6 and 7.

Table 1 provides the nucleotide sequence and deduced amino acid sequence of a genomic clone isolated from a lambda gtll cDNA library prepared from human U-937 monocytic cells stimulated with the mitogen

phorbolmyristate acetate (PMA). This clone encodes a 117-amino acid polypeptide that is 19.82% identical and 48.65% similar to human TNF.

A polymerase chain reaction (PCR) product

corresponding to human HSF is shown in Table 3. This product was obtained by amplifying DNA from human TYH histiocytic cells using primers designed based on rabbit HSF peptide sequences, as described in Example 5.

Vectors and Host Cells

The nucleotide sequences of the present invention can be incorporated into a DNA vector for propagation and/or expression in a host cell. Such vectors typically contain a replication sequence capable of effecting replication of the vector in a suitable host cell as well as sequences encoding a selectable marker, such as an antibiotic resistance gene. Upon transformation of a suitable host, the vector can replicate and function independently of the host genome or integrate into the host genome. Accordingly, the present invention provides a vector comprising a nucleotide sequence of the present invention and a replication sequence.

In addition, the present invention provides an HSF expression vector comprising a nucleotide sequence encoding HSF and control sequences capable of effecting expression of HSF in a suitable host cell. The control sequences include sequences that facilitate transcription of HSF DNA and translation of HSF mRNA, such as a

promoter, an optional operator and/or enhancer, a

ribosome binding site, and termination sequences. The HSF expression vector can also comprise a suitable replication sequence and a selectable marker. In

addition, the HSF expression vector can comprise an amplifiable gene (e.g., the dihydrofolate reductase gene) which allows selection of host cells containing multiple copies of the nucleotide sequence encoding HSF.

The present invention also provides host cells transformed or transfected with a vector of this

invention. A wide variety of eukaryotic and prokaryotic host cells are available for propagation and/or

expression of the constructs of the present invention.

Examples include mammalian cells, yeasts and other fungi, plant cells, and phage. Vector-host systems are well-known, and thus the selection or design of an appropriate vector for propagation and/or expression in a given host is within the level of skill in the art. The vectors of the present invention are introduced into a selected host by any convenient method, which will vary depending on the vector-host system employed. HSF Polypeptides and Analogs

The present invention provides an isolated HSF polypeptide or HSF analog as well as a peptide derived therefrom. The HSF polypeptide or analog is

substantially free from TNF and includes the polypeptides described above. In one embodiment, the HSF polypeptide is substantially pure as determined by immunoblotting and silver staining.

HSF polypeptides can be purified from any convenient source, including serum (as exemplified in Example 1) and the culture medium of cells that secrete HSF, such as histiocytes (as exemplified in Example 2).

Alternatively, HSF polypeptides or analogs can be

obtained using standard techniques to express a DNA sequence encoding HSF and to purify the resultant

recombinant protein. HSF polypeptides or analogs can also be produced synthetically.

The HSF polypeptides or analogs of the present invention are useful as anti-tumor agents, particularly because, unlike TNF, HSF is not cytotoxic. Accordingly, HSF polypeptides and analogs can be used as cancer

chemotherapeutic agents and as research tools in

determining the biochemical basis of tumor regression.

Furthermore, the HSF polypeptides of the present

invention are also useful as molecular weight markers.

The present invention also provides a peptide

corresponding to a portion of the HSF polypeptide or analog of the present invention. In one embodiment, this peptide comprises at least 15 N-terminal amino acids, and in a variation of this embodiment, at least 25 amino acids of the HSF polypeptide or analog. Conveniently, the peptide includes an amino acid sequence unique to the HSF polypeptide or analog, and more conveniently, the peptide consists exclusively of a such a unique sequence. Unique sequences are readily identified by conducting a computer-based homology search.

Peptides can be produced by enzymatic or chemical cleavage (as described in Example 3), recombinant technology, or chemical synthesis. Such peptides are useful for inducing anti-HSF peptide antibodies. In addition, any peptide containing an epitope in common with a non-denatured HSF polypeptide or analog can be used in determining the specificity of antibody

compositions and in purifying anti-HSF antibody

compositions. Furthermore, such peptides are also useful as controls in assays and as analyte analog in

competitive assays.

Anti-HSF Antibodies

The present invention also provides an anti-HSF antibody. The anti-HSF antibody binds the HSF

polypeptide, analog, or peptide of the present invention (hereinafter "HSF-based antigen") and exhibits

insubstantial binding to other cytokines, such as TNF and IL-1. The antibody generally has an affinity for the

HSF-based antigen sufficient to allow its detection in an immunoassay in the presence of one or more other

cytokines. The antibody affinity required for detection of an HSF-based antigen using a particular immunoassay method will not differ from that required to detect other polypeptide antigens.

The anti-HSF antibody can be induced using any of the HSF-related antigens described above. Peptides used for production of anti-HSF antibodies typically share not more than about 40% homology, conveniently not more than about 20% homology, and more conveniently 10% homology or less, with other cytokines. In one variation of this embodiment, a peptide used for production of a specific anti-HSF antibodies consists exclusively of a sequence unique to the particular HSF-related antigen. The anti- HSF antibody can be polyclonal or monoclonal.

Polyclonal anti-HSF antibodies are induced by administering an immunogenic composition comprising an HSF-related antigen of this invention to a host animal. Preparations of immunogenic compositions may vary

depending on the host animal and are well known. For example, peptides may be conjugated to an immunogenic substance such as KLH or BSA or provided in an adjuvant or the like.

The induced antibody composition can be tested to determine whether the composition is specific for the HSF-related antigen. If a polyclonal antibody

composition does not provide the desired specificity, the antibodies can be purified to provide the desired

specificity by a variety of conventional methods. For example, the composition can be contacted with one or more other cytokines affixed to a solid substrate to remove those antibodies that cross-react with the other cytokines.

The composition can also be purified to reduce binding to other substances by contacting the composition with the HSF-related antigen affixed to a solid substrate so that those antibodies that bind to the antigen are retained. Purification techniques using peptides or antibodies affixed to a variety of solid substrates such as affinity chromatography materials, including Sephadex, Sepharose, and the like are well known. Exemplary preparations of polyclonal anti-HSF peptide antibodies are described in Example 3.

Monoclonal anti-HSF antibodies can also be prepared by conventional methods. Briefly, a mouse is injected with an immunogenic composition comprising an HSF-related antigen of this invention and spleen cells obtained.

These spleen cells are fused with a fusion partner to prepare hybridomas. Antibodies secreted by the

hybridomas are then typically screened to select a hybridoma secreting an anti-HSF antibody of the desired specificity. Hybridomas that produce the desired antibody are cultured by standard techniques. Hybridoma preparation techniques and culture methods are well known and constitute no part of the present invention.

This invention is further illustrated by the following specific, but non-limiting, examples.

Procedures that are constructively reduced to practice are described in the present tense, and procedures that have been carried out in the laboratory are set forth in the past tense.

EXAMPLE 1

Purification and Characterization of Rabbit HSF

Preparation of Stimulated Rabbit Serum

Rabbit HSF was purified from the serum of rabbits inoculated with a Propionibacterium acnes vaccine. The vaccine was prepared as follows: P. acnes IID 912 was cultured in GAM broth (Nissui Pharmaceutics, Tokyo,

Japan) under anaerobic conditions at 37°C for 3 days.

Bacteria were killed in 1% formalin and washed 3 times in physiological saline.

One hundred mg/rabbit of P. acnes vaccine was

administered intravenously to Japanese Albino rabbits, weighing 2.3-2.5 kg (Kanamaru Dobutu, Tokyo, Japan). Nine days later, the rabbits were injected with 100 mg/ rabbit of lipopolysaccharide (LPS) from Escherichia coli O111:B4w (Difco, MI, USA). Blood was collected 90 min. after LPS administration, and serum (hereinafter " stimulated serum") was obtained by venopuncture and followed by centrifugation.

Purification of HSF was carried out by monitoring in vivo anti-tumor activity and in vitro cytotoxicity. In vivo anti-tumor activity was evaluated using transplanted murine Meth A sarcomas. Briefly, approximately 5 X 10⁵ Meth A sarcoma cells were injected subcutaneously into Balb/c mice. When tumors had grown to 7-8 mm in

diameter, 0.1 ml/tumor of test samples was injected into the tumors.

The tumor diameter was monitored, and the relative tumor volume was calculated using the formula described in Haranaka et al., 34 Int. J. Cancer 263-267 (1984).

In vitro cytotoxicity was determined by the same protocol as that for measuring TNF cytotoxicity in vitro. L cells (mouse fibroblast cells), which are sensitive to TNF, were assayed for cytotoxicity by dye exclusion as described in Haranaka et al., 50 Brit. J. Cancer 471-479 (1984). In the purification protocol described below, test samples that exhibited in vivo anti-tumor activity with minimal cytotoxicity in the dye exclusion assay were selected for further purification.

Purification of HSF

Ammonium sulfate was added to stimulated serum to a final concentration of 7% and the mixture stored at 4°C for over 2hrs. The resulting precipitate was collected by centrifugation at 2500 rpm for 20min at 4 °C.

0.5 gm precipitate was then dissolved in 5 ml 40 mM Tris buffer (pH 7.8) and subjected to Q Sepharose Fast Flow column chromatography over a 2.6 X 100 cm column ( Pharmacia Fine Chemicals, Uppsala, Sweden).

Sample was eluted using a linear NaCl gradient of from 0.15 M to 1.00 M in 40 mM Tris buffer (pH 7.8).

Anti-tumor activity eluted at high (above 200 mM) NaCl concentration. Fractions eluting at 200-400 mM NaCl and exhibiting cytotoxicity in addition to anti-tumor activity were confirmed to contain TNF by neutralization of bioactivity using an anti-rabbit TNF antibody.

Fractions eluting at approximately 700 mM NaCl exhibited anti-tumor activity without cytotoxicity. The elution pattern on Q Sepharose chromatography is shown in FIG. 1A. FIG. 1B shows the anti-tumor activity of

HSF-containing Q Sepharose fractions 37 and 39.

Fractions 37-40 were concentrated by ultrafiltration using an Amicon (Beverly, MA) ultrafilter for further purification.

Concentrated Q Sepharose eluate was chromato-graphed over a 2.5 X 50 cm Phenyl Sepharose CL 4B column (

Pharmacia Fine Chemicals, Uppsala, Sweden). Sample was eluted using a linear ammonium sulphate gradient of from 0.80 M to 0.00 M in phosphate buffered saline (PBS; pH 7.2) . Fractions exhibiting anti-tumor activity without cytotoxicity (hereinafter "active fractions") eluted at 0.5-0.7 M ammonium sulfate.

Active fractions were pooled and further purified via high performance liquid chromatography (HPLC) using a Shimadzu LC 6A HPLC (Shimadzu, Kyoto, Japan). Sample was loaded onto a 0.75 X 7.5 cm hydrophobic TSK-gel Phenyl 5PW column (Toyosoda, Tokyo, Japan) and eluted using a linear ammonium sulfate gradient of from 1.00 M to 0.00 M in PBS (pH 7.2). Active fractions eluted at 0.5 M

ammonium sulfate. Active fractions were pooled and subjected to HPLC on a 0.75 X 15 cm Shimpack Diol 150 gel filtration column (Shimadzu, Kyoto, Japan). Sample was eluted with PBS (pH 7.2).

The eluent from the Diol 150 HPLC was further purified via HPLC over a 0.75 X 15 cm Shimpack DEAE 5PW anion exchange column (Shimadzu, Kyoto, Japan). HSF was eluted with a linear NaCl gradient of from 0.00 M to 1.00 M in 40 mM Tris buffer (pH 7.8).

Molecular Weight Determinations

The molecular weight (MW) of purified rabbit HSF in the non-denatured state was determined by gel filtration over the Diol 150 HPLC column described above, eluting with PBS (pH 7.2) at a constant flow rate of 1 ml/min. 0.5 ml fractions were collected. MW was determined by comparison of the elution pattern with that observed for the components of a low MW gel filtration calibration kit (Pharmacia Fine Chemicals, Uppsala, Sweden; kit contained ribonuclease A, 13,700 kilodaltons (kD); chymotrypsinogen A, 25,000 kD; ovalbumin, 43,000 kD; and bovine serum albumin, 67,000 kD). The MW of non-denatured rabbit HSF was found to be about 45-51 KD.

The MW of purified rabbit HSF in the denatured state was determined by gel electrophoresis according to the method of Laemmli. Briefly, purified rabbit HSF was electrophoresed with MW standards on a 12% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE), which was then silver stained. The MW standards employed were from a commercially available electrophoresis kit (Bio Rad, CA, USA; kit contained lysozyme, 14,400 kD; trypsin

inhibitor, 20,100 kD; carbonic anhydrase, 29,000 kD;

ovalbumin, 45,000 kD; bovine serum albumin, 67,000 kD; and phosphorylase, 92,500 kD). Three bands were observed at approximately 35, 42, and 55 kD.

Amino Acid Sequencing of HSF Peptides

Rabbit HSF was purified as described in Example 1, and the absence of TNF from the preparation was

determined by immunoblotting using anti-rabbit TNF.

Purity was confirmed by silver staining. Peptide were generated from purified rabbit HSF by enzymatic digestion and then sequenced. First,urea was added to 0.5 ml purified rabbit HSF to a final concentration of 4M (pH 8.0). This preparation was treated with acromobacter protease (Seikagakusya, Tokyo, Japan) at an enzyme:

substrate ratio of 1:100 for 24 hrs. at 37°C.

The resultant peptides were purified by reverse phase

HPLC using a Hitachi HPLC 650 (Hitachi, Tokyo, Japan) on a 0.25 X 17.5 cm SMB-ODS column (Shimadzu, Kyoto, Japan) with a linear acetonitrile gradient of from 0% to 72% in the presence of 0.1% trifluoroacetate. The elution pattern is shown in FIG. 2.

Sequencing was performed with a Hitachi HPLC 650, Sequencer 470A, Amino Acid Analyzer 120A (Hitachi, Tokyo, Japan).

Fractions corresponding to the ten main peak fractions of enzyme-digested HSF (see FIG. 2) were collected and sequenced. Three portions of the amino acid sequence were determined from among the 10 peaks. Sequencing of these peaks allowed identification of four portions of the amino acid sequence:

Peak 3 - PNAYL SEQ ID NO.1;

Peak 4 - LPPGLLAPMRQLRS SEQ ID NO.2;

Peak 5 - NLEXFTNGMEQHYAQL SEQ ID NO.3;

Peak 8 - NPAENQAHELPNQLN SEQ ID NO.4. A computer-based homology search demonstrated that these sequences were novel.

Anti-tumor Activity and Cytotoxicity of Rabbit HSF

The in vivo anti-tumor activity of partially purified rabbit HSF was compared with that of TNF in the murine Meth A sarcoma model described above.

TNF induced necrosis (mostly central necrosis) of transplanted tumors at an early stage after its

administration (1-2 days). When a high dose of TNF was administered, complete regression of the tumors was achieved. At lower doses, regrowth of the tumors from the periphery of the area of tumor necrosis was observed.

In contrast, HSF exhibited suppressive effects on tumor growth but no necrotizing activity against the transplanted murine tumors. The tumor sizes remained constant or decreased slightly at one week after a single injection of HSF intratumorly.

Statistically significant differences between the control and HSF samples were evident at every

purification step.

In vitro cytotoxicity studies revealed large

differences between TNF and HSF. TNF was strongly

cytotoxic to L cells, as well as several other cell lines. In contrast, HSF displayed no cytotoxicity to L cells.

EXAMPLE 2

Purification and Characterization of Human HSF

Human HSF was purified from TYH cells, a histiocytic cell line established from the peripheral blood of a patient with malignant lymphoma (Haranaka et al., 36 Int. J. Cancer 313-319 (1985)). TYH cells were cultured in high GEM medium (Flow Lab., VA) supplemnted with 10% fetal calf serum. At confluence, the cultures were diluted ten-fold with Ham's F12 medium (Flow Lab., VA), and the cells were cultured an additional three days. Cells were removed by filtration using a GF 10 Whatmann glass filter. The filtrate was subjected to ultrafiltration using the Pericon Cassette system ( Millipore Corp. Bedford, MA). Human HSF was then purified from the concentrate by the same procedure described for rabbit HSF in Example 1.

The MW of human HSF in the non-denatured state was determined by Diol 150 HPLC as described in Example 1. HSF activity eluted at around 10 min., indicating a MW of about 45-51 kD.

Amino acid sequencing of purified human HSF was attempted, but the N-terminus was blocked.

EXAMPLE 3

Production of Anti-HSF Peptide Antibodies To prepare polyclonal anti-HSF peptide antibodies, synthetic peptides corresponding to the following partial rabbit HSF amino acid sequences were prepared:

7-mer: PMRQLRS SEQ ID NO.5;

13-mer: NPAENQAHELPNQ SEQ ID NO.6. A computer-based homology search on these sequences produced the alignments shown in Table 4 (7-mer) and Table 5 (13-mer).

The proteins listed in Tables 4 and 5, in addition to rabbit and human HSF are as follows: IL-1, human interleukin-lα; IL-lrec, rat IL-1 receptor (type 1) ; IL- lrec2, murine IL-1 receptor (type2) ; IL-lrecag, human IL-1 receptor antagonist; TNF, human tumor necrosis factor-α; HPTP, human protein-tyrosine phosphatase epsilon precursor.

DDY mice (Krea Japan, Tokyo, Japan) were immunized with either the 7-mer or the 13-mer conjugated to " multiple antigen peptide" (MAP) in complete Freund's adjuvant (Iatorn, Tokyo, Japan) by injection into the foot pad. Five additional rounds of immunization were carried out by intraperitoneal injection of

MAP-conjugated peptides in incomplete Freund's adjuvant (Iatorn, Tokyo, Japan). Whole blood was then collected from the postorbital venous plexus. Anti-peptide serum was isolated and stored frozen in small aliquots.

The resultant polyclonal antibodies were designated anti-7-mer and anti-13-mer, respectively. EXAMPLE 4

Immunoblotting of Human and Rabbit HSF Human and rabbit HSF were analyzed by

immunoblotting. Partially purified human and rabbit HSF were electrophoresed on a 12% SDS-polyacrylamide gel, and the proteins were transblotted onto a membrane.

Different sections of the membrane were reacted with anti-7-mer antibody and anti-13-mer antibody. Primary antibody binding was visualized using alkaline

phosphatase-conjugated goat anti-mouse antibody (

Stratagene, LaJolla, CA). Color development on the filter was done using BCIP-NBT (5-Brom-4-chloro-3-indolyl

phosphate-nitro blue tetrazolium; Stratagene, LaJolla, CA) according to the manufacturer's instructuiins. FIG. 3A shows that the roughly the same pattern of bands is observed regardless of the anti-peptide antibody used. In the human HSF samples, the anti-peptide antibodies reveal two proteins of approximately 41 and 46 kD. In the rabbit HSF samples, the antibodies react with three proteins of approximately 35, 42, and 55 kD.

EXAMPLE 5

Polymerase Chain Reaction Amplification of Human HSF

To amplify human HSF gene sequences from TYH cells using the polymerase chain reaction (PCR), a number of oligonucleotide primers were designed based on the rabbit HSF peptide sequences. Primers were synthesized using an ABI394 Oligonucleotide Synthesizer (Applied Biosystems Inc. Foster City, CA)

Template DNA from TYH cells was prepared by the crude cell lysate procedure, according to Davis, L.G. et al., Basic Methods in Molecular Biology (Elsevier, 1986).

DNA amplification was performed using Thermus aαuaticus (Tag) polymerase in a Perkin-Elmer Corp.

Norwalk, CT) DNA thermal cycler. Template DNA was combined with 5' primer, 3' primer, dNTP mix, and Tag polymerase in PCR reaction buffer (500 mM KCl, 10 mM

Tris-HCl (pH 8.3), 25 mM MgCl₂, 20 mg/ml gelatin) in polypropylene microcentrifuge tubes. These reaction mixtures were overlaid with mineral oil to prevent evaporation during amplification.

The reaction mixtures were denatured at 94°C for a 1-min. period, annealed at 42°C for 2 min., and extended with Tag polymerase at 72°C for 3 min. through 35 cycles of PCR. The reaction mixtures were held at 72°C for 10 min. and then stored at 4°C. A 0.3 kb PCR product was obtained when the following primers were used to amplify TYH DNA: 5' primer: AGGTACGGCAACCACTT SEQ ID NO. 7; 3' primer: GAGAACCAGGCTCACGA SEQ ID NO. 8. The same primers were used to amplify DNA from U-937 (human histiocytic lymphoma) cells, which revealed positive PCR products of 0.7 kb in length. In addition, amplification of a cDNA library derived from U-937 cells stimulated with 50 ng/ml phorbolmyristate acetate (PMA) for 3-1/2 days also yielded positive PCR products of 0.8 kb in length.

TYH PCR products were purified using Prime Erase Quick Column (Stratagene, LaJolla, CA) according to the manufacturer's instructions. The nucleotide sequences of the TYH products were then determined by direct

sequencing, as described by Davis, L.G. et al., Basic Methods in Molecular Biology (Elsevier, 1986), using Tag polymerase-catalyzed amplification reactions and Applied Biosystems Model 373A DNA sequencer (Applied Biosystems, Foster City, CA). The sequences obtained are listed in Table 3 (SEQ ID NO. 9).

EXAMPLE 6

Immunoscreening of Human U-937 cDNA Library A cDNA library was constructed in the lambda gtll bacteriophage expression cloning vector developed by

Richard Young and Ronald Davis (80 Proc. Natl. Acad.

Sci., USA 1194-1198 (1983) according to a modified Gubler & Hoffman procedure (25 Gene 263-269 (1983)).

cDNA inserts are cloned into the carboxyl end of the β-galactosidase gene-coding region (lac Z) in lambda gtll. This vector is designed to express the cDNA insert as a β-gal-cDNA fusion protein. One of every six recombinants is expected to be in the correct reading frame and

orientation for appropriate translation. Expression of fusion proteins is triggered by induction of the lac Z gene with isopropyl-β-D-thiogalacto-pyranoside (IPTG). The library is then screened with an antibody against the protein of interest.

Here, the library was prepared using poly-adenylated mRNA from U-937 cells treated with PMA at 50 ng/ml for 3-1/2 days to induce differentiation to a monocyte-like stage. The mRNA was primed with an oligo(dT) primer containing an Eco RI restriction enzyme site for cloning into lambda gtll. cDNA was produced using reverse transcriptase. cDNAs less than about 1kb were removed, and the remaining cDNA, which ranged from about 1.0 to 3.9 kb were cloned into the lambda gtll vector.

The resultant library was amplified once and yielded 1.2 X 106 independent plaques with the host bacterial strain E. coli Y1090r-.

The library was screened using the picoBLUE

immunoscreening kit (Stratagene, LaJolla, CA) which is designed for screening phage plaques blotted onto

nitrocellulose filters (Huynuh, T.V., et al, "DNA

cloning: A practical approach", Vol. 1, pp.49-78. (D.M. Glover ed., IRL Press, Oxford)). Plaque screening was done according to the manufacturer's instructions.

Briefly, the bacteria were combined with the phage and incubated for 15 min. at 37°C. Top agar, at 37°C, was added, and after mixing, the mixture was poured onto agar plates, which were then incubated at 42°C for 3.5 hours. When small plaques became visible, an IPTG-treated nitrocellulose filter was applied to each plate, and the plates incubated at 37°C for 3.5 hours. The nitrocellulose filters were then removed.

To detect recombinants containing HSF DNA sequences, the filters were probed with a mixture of anti-7mer and anti-13mer mouse antibodies (primary antibodies),

produced as described above in Example 3. Immunoscreening was performed according to Clontech's instruction.

Primary antibody binding was visualized using alkaline phosphatase-conjugated goat anti-mouse antibody (

Stratagene, LaJolla, CA). Color development on the filter was done using BCIP-NBT. Positive clones were picked and subjected to three additional rounds of plating and immunoscreening.

Positive clones were further screened by PCR as described by Saiki, R.K. , et al. (230 Science 1350-135 (1985)). Lambda DNA was obtained by inserting a toothpick into positive plaque from the culture plates and transferring the toothpick to distilled water. The insert DNA in each preparation was amplified using a lambda gtll" cDNA insert screening amplimer set (

Clonotech, Palo Alto, CA). This kit provides lambda gtll amplimer (primer) sequences that bind to sites flanking the EcoRI cloning site. The sequences of these primers are:

5' primer: GACTCCTGGAGCCCG

3' primer: GGTAGCGACCGGCGC

PCR reaction mixtures were set up as described above in Example 5, and 35 cycles of amplification were performed as follows. The reaction mixtures were denatured at 94°C for a 1-min. period, annealed at 60°C for 1 min., and extended with Tag polymerase at 72°C for 3 min. through 35 cycles of PCR. The reaction mixtures were held at 72°C for 10 min. and then stored at 4°C.

Three positive clones of different lengths were selected for further analysis and the corresponding PCR products purified using the Prime Erase Quick Column (Stratagene LaJolla, CA). EXAMPLE 7

Sequencing of Positive Clones from U-937 cDNA Library The nucleotide sequences of PCR products

corresponding to three clones from the U-937 library ( isolated as described in Example 6) were determined by direct sequencing using Tag polymerase-catalyzed

amplification reactions and Applied Biosystems Model 373A DNA sequencer (Applied Biosystems, Foster City, CA) as described above in Example 5. The sequences obtained where then compared with those in the Swiss-prot protein sequence database.

The nucleotide sequence of a 934-base pair (bp) clone was determined initially by direct PCR product sequencing and then confirmed with double-stranded sequencing using primer walking methods. This sequence is set forth in Table 1 (SEQ ID NO. 10). This sequence was determined to be novel sequences by computer-based homology search.

The deduced amino acid sequence for the 934-bp HSF clone is shown in Table 1 (SEQ ID NO. 11). A potential glycosylation site is indicated in Table 1 by an

asterisk. The 934-bp HSF clone has an open reading frame of 117 amino acids interrupted by a non-coding region. This non-coding region was determined to be an intron based on comparison of the amino acid sequences of the human HSF clone to the amino acid sequences of the rabbit HSF peptide. An alignment of these sequences is set forth in Table 6. Thus, the 934-bp clone was derived from genomic DNA contaminating the U-937 cDNA library.

Table 6

Alignment of Rabbit Peptide Sequences with the HSF Amino Acid Sequence Deduced from the 934-bp Human HSF Clone

The other positive clone was identified as human protein-tyrosine phosphatase epsilon precursor and was therefore not further analyzed.

The predicted locations of a number of restriction endonuclease recognition sites in the 934-bp HSF clone are set forth in Table 7. Approximate locations of restriction sites are provided in terms of the number of nucleotides 3' to the EcoRI site at 5' end of the insert.

A sequence comparison of the deduced amino acid sequence of the 934-bp HSF clone (hereinafter "the human HSF amino acid sequence") with the amino acid sequence of human TNF indicated 19.82% identity and 48.65% similarity (Table 8).

In addition, analysis of this amino acid sequence using the PROSITE data base (Bairoch, A., Switzerland)

identified a number of sequence motifs which are set forth in Table 9.

The predicted locations of a number of enzymatic and chemical cleavage sites are set forth in Table 10. Approximate locations of cleavage sites are numbered from the N-terminal of methione.

Table 11 shows a several predicted characteristics, such as molecular weight and isoelectric point, along with a summary of the amino acid make-up of human HSF.

Additional information was obtained from various computer programs that predict protein structure based on primary sequence as described below. The HSF amino acid sequence was analysed using PEPPLOT program (obtained as part of the Genetics Computer Group Sequence Analysis Software Pakage, Version 7.3.1-Unix, September 1993 from Genetics Computer Group, Madison, Wisconsin). This program analyzes the following features of protein structure: charged-polar-hydrophobicity of amino acid residues; tendency to form or break beta pleated sheet; Chou-Fasman-predicted tendency to form alpha helices or beta-plated sheets; Chou-Fasman-predicted N-terminal end; Chou-Fasman-predicted C-terminal end; helical

hydrophobic moment; and hydropathy and hydrophilicity. The results of this analysis are shown in Fig. 4.

The HSF amino acid sequence was also analyzed using the PEPTIDESTRUCTURE program (obtained as part of the

Genetics Computer Group Sequence Analysis Software

Pakage). This program predicts secondary structural features, including alpha helices; beta-pleated sheets; colis; and turns. The program also provides predicted values for antigenicity, flexicibility, hydrophobicity, and surface probability. The results of the

PEPTIDESTRUCTUKE analysis of the HSF amino acid sequence are set forth in Table 12.

Helical wheel projections of alpha helices in HSF are shown in Fig. 5. Hydrophobic amino acid residues are boxed.

Claims

WHAT IS CLAIMED IS:

I. An isolated nucleotide sequence encoding, or

complementary to a sequence encoding, histiocyte- secreted factor or an analog thereof.

II. The isolated nucleotide sequence of Claim 1 wherein the nucleotide sequence is a DNA sequence. III. The isolated nucleotide sequence of Claim 2 wherein the nucleotide sequence is a cDNA sequence.

IV. The isolated nucleotide sequence of Claim 1 wherein the nucleotide sequence encodes human histiocyte- secreted factor.

V. An isolated nucleotide sequence selected from the group consisting of:

A. a nucleotide sequence that encodes a

polypeptide comprising an amino acid sequence as set forth in Table 1;

B. a nucleotide sequence complementary to the nucleotide sequence defined in (a); and

C. a nucleotide sequence that hybridizes with one of the nucleotide sequences defined in

(a) and (b) under stringent conditions. VI. The isolated nucleotide sequence of Claim 5 selected from the group consisting of:

A. a nucleotide sequence that encodes a

polypeptide consisting essentially of the amino acid sequence set forth in Table 1; B. a nucleotide sequence complementary to the nucleotide sequences defined in (a); and C. a nucleotide sequence that hybridizes with one of the nucleotide sequences defined in (a) and (b) under stringent conditions. VII. The isolated nucleotide sequence of Claim 6 selected from the group consisting of:

A. a nucleotide sequence that encodes a

polypeptide having the amino acid sequence set forth in Table 1;

B. a nucleotide sequence complementary to the nucleotide sequence defined in (a); and C. a nucleotide sequence that hybridizes with one of the nucleotide sequences defined in (a) and (b) under stringent conditions.

VIII. The isolated nucleotide sequence of Claim 1 wherein the nucleotide sequence is a naturally occurring histiocyte-secreted factor nucleotide sequence. IX. The isolated nucleotide sequence of Claim 1 selected from the group consisting of:

A. a nucleotide sequence as set forth in

Table 1;

C. a nucleotide sequence that hybridizes with one of the nucleotide sequences defined in (a) and (b) under stringent conditions. X. The isolated nucleotide sequence of Claim 1 wherein the nucleotide sequence encodes a histiocyte- secreted factor analog comprising a mutation

consisting of a conservative amino acid

substitution. XI. The isolated nucleotide sequence of Claim 10 wherein said conservative amino acid substitution consists of the replacement of a native amino acid from a selected amino acid group with a different amino acid from the same amino acid group, wherein said amino acid group is selected from the following:

A. K, R, and H;

B. D and E;

(c) N and Q;

(d) F, Y, and W;

(e) P, G, A, V, L, I, and M; and

(f) S and T. XII. An isolated nucleotide sequence comprising a portion of a histiocyte-secreted factor genomic sequence.

XIII. A DNA vector comprising the nucleotide sequence of Claim 1, and an element selected from the group consisting of:

(a) a replication sequence capable of effecting replication of said DNA vector in a

suitable host cell; and

(b) control sequences capable of effecting

expression of said nucleotide sequence of

Claim 1 in a suitable host cell.

XIV. A host cell transformed with the DNA vector of Claim 13.

XV. An isolated polypeptide selected from the group

consisting of a histiocyte-secreted factor

polypeptide or an analog thereof. XVI. The isolated polypeptide of Claim 15, wherein said polypeptide is human histiocyte-secreted factor.

XVII. An isolated polypeptide comprising an amino acid sequence as set forth in Table 1.

XVIII. The isolated polypeptide of Claim 17

consisting essentially of the amino acid sequence set forth in Table 1.

XIX. The isolated polypeptide of Claim 18 having the amino acid sequence set forth in Table 1.

XX. The isolated polypeptide of Claim 15 wherein the polypeptide is a histiocyte-secreted factor analog comprising a mutation consisting of a conservative amino acid substitution.

XXI. The analog of Claim 21 wherein said conservative amino acid substitution consists of the replacement of a native amino acid from a selected amino acid group with a different amino acid from the same amino acid group, wherein said ammo acid group is selected from the following:

A. K, R, and H;

B. D and E;

(c) N and Q;

(d) F, Y, and W;

(e) P, G, A, V, L, I, and M; and

(f) s and T.

XXII. A anti-histiocyte-secreted factor antibody.