Murine Interleukin-5 Receptor

1. Field of the Invention

The present invention relates to isolated cDNA sequences coding for murine interleukin 5 receptors, murine secretory interleukin 5 receptors and human interleukin 5 recepters and to murine interleukin 5 receptors, murine secretory interleukin 5 receptors and human interleukin 5 receptors which are produced using the isolated cDNA sequences as well as to methods of producing the interleukin 5 receptors.

2. Prior Art

Interleukin 5 (referred to as "IL-5", hereinafter) is a proliferation and differentiation factor for eosinophils and B lineage cells (Immunol. Rev. 102: 29, 107.,1988). It has been known that IL-5 is produced especially by T cells primed with Mycobacterium tuberculosis, parasites or allo-antigens (J. Immunol. 140: 1175, 1988; Nature, 324: 70, 1986). IL-5 has also been known to induce production of IgM class immunoglobulin including anti-DNA antibody. Recently, IL-5 has been suspected of involvement in autoimmune diseases and there is a report that IL-5 is closely associated with eosinophilia accompanied by autoantibody production, fascitis and myositis (Eosinophils, Oxford University Press, 1988).

There are two types of IL-5 receptors (referred to as "IL-5R", hereinafter), namely, membrane bound IL-5R and secretory IL-5R. Among them, mouse secretory IL-5R is able to bind to human IL-5 and therefore expected to serve as a therapeutic agent for diseases associated with IL-5.

The inventors have obtained IL-5 responsive early B cells, T 88 and T-88M by culturing mouse bone marrow cells in the presence of IL-5 (Growth Factors 1: 135 1989) and produced IL-5R. The cross-linking reaction and subsequent SDS-PAGE analysis have revealed that IL-5R comprises at least two types of subunits, one having a molecular weight of about 46,500 and the other having a molecular weight of about 114,000, and that there are two types of IL-5R, a low affinity IL-5R having the dissociation constant of 27 nM and a high affinity IL-5R having the dissociation constant of 150 pM. It has been believed that the low affinity IL-5R comprises the small subunit of an estimated molecular weight of 46,500 while a high affinity IL-5R comprises the large subunit (MW: 114,000) and the small subunit (46,500) (Proc. Natl. Acad. Sci. USA 86: 2311, 1989).

The inventors have produced H7 and T21 monoclonal antibodies by immunizing rats with a membrane fraction of T88-M cells, which inhibit the binding of IL-5 to IL-5R (Int. Immunol. 2: 181, 1990; J. Immunol. 144: 4218, 1990). Anti-IL-5R antibodies, H7 and T21, are found to bind to glycoprotein of the molecular weight of about 60,000 according to the SDS-PAGE analysis. The real molecular weight of the small subunit is found to be about 55,000 according to the binding assay using IL-5 free of an oligosaccharide, suggesting that the low affinity IL-5R comprises a single molecule of molecular weight of about 60,000 (Int. Immunol. 2: 181, 1990).

We have also reported recently that IL-5R is found on the cell surface of human eosinophils. The dissociation constant of human IL-5R is 170-330 pM and the molecular weight is 55,000-60,000 according to the SDS-PAGE analysis. Human IL-5R appeares to be comparable to a low affinity murine IL-5R (Migita, M., Yamaguchi, N., Mita, S., Higuchi, S., Hitoshi, Y., Yoshida, Y., Tomonaga, M., Matsuda, I., Tominaga, A., Takatsu, K., 1991, Cellular Immunology, 133: 484-497).

There has been no report on the isolation of a DNA sequence coding for the low affinity murine/human IL-5R. An object of the invention is to isolate the DNA sequence coding for the low affinity murine/human IL-5R and to determine the DNA sequence. The isolated DNA sequence may be used to produce murine/human IL-5R in mammalian cells. Another object of the invention is to obtain a DNA sequence coding for secretory IL-5R which is distinct from the DNA sequence coding for membrane bound IL-5R and to produce pure secretory IL-5R using the DNA.

The present invention is characterized by the following description:

(1). An isolated cDNA sequence coding for murine interleukin 5 receptor which is synthesized from murine early B cell mRNA.

(2). The isolated cDNA sequence of (1) wherein the nucleotide sequence comprises the open reading frame sequence described in SEQ ID No. 1.

(3) The isolated cDNA sequence of (1) wherein the nucleotide sequence comprises the entire sequence described in SEQ ID No. 2.

(4). An isolated cDNA sequence coding for secretory murine IL-5R which is synthesized from murine early B cell mRNA.

(5) The isolated cDNA sequence of (4) wherein the nucleotide sequence comprises the open reading frame sequence described in SEQ ID No. 3.

(6). The isolated cDNA sequence of (4) wherein the nucleotide sequence comprises the entire sequence described in SEQ ID No. 4.

(7). An isolated murine interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID No. 5.

(8). An isolated murine interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID No. 6.

(9) An isolated murine secretory interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID No. 7.

(10) An isolated murine secretory interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID No. 8.

(11). A method of producing the murine interleukin 5 receptors which comprises culturing cells capable of expressing the murine interleukin 5 receptors in medium and isolating the murine interleukin 5 receptors from the cells or the culture supernatant using anti-interleukin 5 receptor antibodies.

(12) A COS 7 monkey cell (ATCC CRL 1651) transfected with a recombinant vector containing the cDNA sequence of any one of (1)-(6).

(13) A method of producing the murine interleukin 5 receptors and the murine secretory interleukin 5 receptors comprises culturing the COS 7 cell transfected with relevant DNA in medium, and recovering the murine interleukin 5 receptors from the cells or secretory murine interleukin 5 receptors from the culture supernatant.

(14). An isolated cDNA sequence coding for human interleukin 5 receptor which is synthesized from mRNA of a human peripheral blood eosinophil.

(15). The isolated cDNA sequence of (14) wherein the nucleotide sequence described in SEQ ID No. 9 comprises the open reading frame sequence coding for human interleukin 5 receptor.

(16). The isolated cDNA sequence of (14) wherein the nucleotide sequence described in SEQ ID No. 10 comprises the entire sequence coding for human interleukin 5 receptor.

(17). The isolated cDNA sequence of (14) wherein the nucleotide sequence described in SEQ ID No. 11 comprises the open reading frame sequence coding for human interleukin 5 receptor 2.

(18). The isolated cDNA sequence of (14) wherein the nucleotide sequence described in SEQ ID No. 12. comprises the entire sequence coding for human interleukin 5 receptor 2.

(19). An isolated human interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID NO. 13.

(20). An isolated human interleukin 5 receptor wherein the amino acid sequence comprises the sequence described in SEQ ID NO. 14.

(21 ). The isolated cDNA sequence of (14) coding for a whole or part of amino acid residue numbers 1-333 described in SEQ ID No. 13

(22). A secretory human interleukin 5 receptor which lacks a cytoplasmic region and a transmembrane region of human interleukin 5 receptor.

(23). An expression vector comprising the cDNA sequence of any one of (14), (15), (16), (17), (18), and (21).

(24). A method of producing the secretory human interleukin 5 receptor and its analogues which comprises culturing a recombinant vector coding for the secretory human interleukin 5 receptor under the conditions which promote the expression thereof and recovering the secretory human interleukin 5 receptor.

The invention provides isolated DNA sequences coding for murine/human IL-5R and pure murine IL-5R produced by a genetic engineering technique using the isolated DNA sequence as well as an isolated DNA sequence coding for secretory murine IL-5R. The DNA sequence coding for secretory murine IL-5R is especially valuable in constructing a nucleotide sequence corresponding to the sequence of secretory human IL-5R and in producing secretory human IL-5R using the DNA sequence. The secretory human IL-5R thus produced may be utilized as a therapeutic agent for autoimmune disorders or diseases with eosinophilia in which IL-5 is believed to be involved and may greatly contribute to the medical and pharmaceutical field.

The Invention is explained referring to the attached drawings.

FIG. 1 shows partial restriction maps of two IL-5R cDNA clones. The box indicates an open reading frame which is expected to be translated. The shaded portion at the 5'-end indicates a signal peptide, and the solid portion indicates the transmembrane region.

FIGS. 2A-2B shows vectors used in the present invention and a site of inserting an isolated DNA into the vector.

FIGS. 3A-3C shows the results of binding assay using .sup.35 S-labeled IL-5 and the Scatchard plot analysis.

FIG. 4 shows the results of cross-linking experiment using .sup.35 S-labeled IL-5.

FIG. 5 shows the results of immunoprecipitation of the translated product of mouse IL-5RcDNA that codes for membrane type IL-5R.

FIG. 6 shows a restriction map of a human IL-5R cDNA fragment of HSIL5R and HSIL5R2. The box represents the open reading frame of IL-5R. The 5' end hatched box is a putative signal peptide and the solid box is the predicted transmembrane region. The mark indicates the starting point of the nucleotide sequence which distinguishes HSIL5R from HSIL5R2.

FIGS. 7A-7D shows Scatchard plot analyses of a binding assay of .sup.35 S-labeled murine IL-5 or .sup.125 I labeled human IL-5 to the pCAGGS.HSIL5R or pCAGGS.HSIL5R2 transfected COS 7 cell. The symbol .quadrature. represents a total binding and the symbol represents a nonspecific binding in the presence of a 100-fold excess of non-radiolabeled IL-5. COS 7 cells were transfected with pCAGGS.HSIL5R and the transformant was tested for binding using .sup.125 I-labeled human IL-5 (FIG. 7A, inset of FIG. 7A). COS 7 cells were transfected with pCAGGS.HSIL5R2 and the transformant was tested for binding using .sup.125 I-labeled human IL-5 (FIG. 7B, inset of FIG. 7B). COS 7 cells were transfected with pCAGGS.HSIL5R and the transformant was tested for binding using .sup.35 S-labeled mouse IL-5 (FIG. 7C, inset of FIG. 7C). COS 7 cells were transfected with pCAGGS.HSIL5R2 and the transformant was tested for binding using .sup.35 S-labeled mouse IL-5 (FIG. 7D, inset of FIG. 7D).

FIG. 8 is a bar graph showing the binding specificity of .sup.125 I labeled human IL-5 to IL-5R. 100 .mu.l of the COS 7 transformants (4.times.10.sup.5 cells) carrying pCAGGS.HSIL5R and 500 pM .sup.125 I-labeled human IL-5 were incubated in the presence of a 1,000-fold excess of cytokines.

FIG. 9 shows a band pattern of chemical cross-linking of IL-5 analysed by SDS-PAGE. COS cells were transfected with a pCAGGS vector and the transformant was incubated with .sup.35 S-labeled murine IL-5 (lane 1) or .sup.125 I-labeled human IL-5 (lane 4). Then a cross-linking agent, bis(sulfosuccinimidyl) suberate was added to the mixture. After incubation, the mixture was electrophoresed under non-reduced condition. Similarly, COS7 cells were transfected with pCAGGS. HSIL5R and the transformants were incubated with .sup.35 S-labeled murine IL-5 in the presence (lane 3) or absence (lane 2) of an excess amount of non-labeled murine IL-5 or with .sup.125 I-labeled human IL-5 in the presence (lane 6) or absence(lane 5) of an excess amount of non-labeled human IL-5. Then, a cross-linking agent, bis(sulfosuccinimidyl) suberate was added to the mixture. After incubation, the mixture was electrophoresed under non-reducing condition.

The description which relates to murine IL-5R is indicated (Murine) and which relates to human IL-5R is indicated (Human).

Preparation of Poly(A).sup.+ RNA from Mouse Bone Marrow Cells (Murine)

In order to prepare the cDNA coding for the IL-5R, mRNAs are recovered from the mouse bone marrow cells having IL-5R. Mouse bone marrow cells are obtainable by a long-term bone marrow cell culture in the presence of IL-5 (Growth Factor 1: 135, 1989). A suitable source of cells may be a Balb/c mouse bone marrow long-term culture cell line, Y16, which is early B cells and shows a strong response to IL-5 (even at a concentration of 1 pg/ml of IL-5). RNA is prepared from the cell according to the method described by Okayama et al. (Methods in Enzymology 154: 3 1987). Poly(A).sup.+ RNA is recovered by fractionating the total RNA with the affinity chromatography using an oligo (dT) cellulose column.

Construction of cDNA Library from mRNA (Murine)

The poly(A).sup.+ RNA is reverse transcribed to cDNA using random primers and reverse transcriptase (Gene 25: 263, 1983). The cDNA larger than 1.0 kb is selected for cloning and inserted into the BstXI site of CDM 8 vector (see FIG. 2A) containing a cytomegalovirus promoter according to the method described by Seed et al. (Proc. Natl. Acad. Sci. USA 84: 8573, 1987). E. coli is transformed with the recombinant plasmid in order to provide cDNA library expressible in mammals.

Cloning of IL-5R Gene: Transfection of COS7 Cells Using the DNA of the Transformant (Murine)

COS 7 cells (Green monkey kidney cells) are transfected with the DNA according to the DEAE dextran or protoplast fusion method. The COS7 transformant is screened using anti-IL-5R antibodies H7 and T21 according to the method described by Seed et al. (Nature 329: 840, 1988). H7 and T21 antibodies and the COS7 suspension are incubated together. After incubation, the mixture is transferred to plates coated with goat anti-rat IgG antibody (H7 and T21 are rat IgG antibodies). Then, plasmid DNA is recovered from the COS7 cells immobilized on the bottom of the plate. The transformation-screening procedure described above is repeated several times. After screening, a group of the selected COS7 transformant is further screened by flow cytometry using fluorescein-conjugated H7 and T21 and the transformant containing IL-5R cDNA is identified.

The Entire Structure of murine IL-5R Genes (Murine)

The rough restriction maps of IL-5R cDNA isolated above are shown in FIG. 1. pIL-5R.8 is the cDNA clone prepared first from the CDM 8 library. pIL-5R.2 is obtained from the cDNA library using the HindIII-PstI fragment of pIL-5R.8 as a probe according to the colony hybridization method.

The nucleotide sequences of the cDNA fragments of pIL-5R.2 and pIL-5R.8 are determined according to the method described by Sanger et al (Proc. Natl. Acad. Sci. USA 74: 5463, 1977). The entire nucleotide sequence of the cDNA fragment of pIL-5R.8 and the deduced amino acid sequence are shown in SEQ ID No. 15. The nucleotide A of the ATG initiation codon is numbered 303 and the amino acid methionine is numbered 1. The cDNA fragment of pIL-5R.8 has 1808 nucleotides in length which codes for 415 amino acids. This polypeptide consists of 4 portions according to Hydropathy plot (OF URFS and ORFS, Rusell F, Doolittle, University Science Books, 1987): singal peptide (See amino acids 1-17 of SEQ ID NO: 15), extracellular region, transmembrane region, and cytoplasmic region. The amino acids at positions 32-34, 128-130, 213-215, 241-243, 392-394, and 412-414 of SEQ ID NO. 15 are presumably linked to N-linked oligosaccharide. pIL-5R.2 lacks a transmembrane region (FIG. 1 and SEQ ID No. 16) and therefore, IL-5R expressed by pIL-5R.2 is a secretory type. As shown in SEQ ID No. 16, pIL-5R.2 lacks the sequence between the nucleotide Nos. 986 and 1164

Expression of IL-5R gene (Murine)

The following two types of vectors, CDM8 and pCAGGS, can be utilized as expression vectors for an isolated IL-5R cDNA sequence (see FIG. 2).

CDM 8 vector: The vector demonstrates an elevated DNA expression in mammalian cells. The vector has two BstXI sites. The vector is digested with BstXI and both ends of cDNA of interest are ligated to a BstXI linker. The cDNA-linker complex is ligated to the BstXI digested vector.

pCAGGS vector: The vector contains a CMV enhancer upstream of the promoter of pAGS-3, which is a vector having a much stronger expression ability than that of CDM8 (Gene, 79:269, 1989). The cDNA insertion site of pCAGGS is XhoI site substituted with EcoRI site in the exon of rabit .beta.-globin gene region. The pCAGGS vector demonstrate a higher level of DNA product expression than pAGS-3.

In the Example of the invention, the pCAGGS vector is used for expression test of IL-5R and the expressed murine IL-5R is tested by IL-5 binding test, IL-5 cross-linking test and immunoprecipitation test using monoclonal antibody, H7.

The murine IL-5R cDNA encoding a secretory IL-5R is inserted into the XhoI (EcoRI) site of the pCAGGS vector. COS7 (Green monkey kidney cell, ATCC CRL1651 ) is transfected with the recombinant plasmid and the resulting transformant is grown in a medium. The amino acid sequence of the peptide in the culture supernatant is determined. The N-terminal 20 amino acids thus determined are the same as those deduced from the nucleotide sequence of the murine IL-5R cDNA. The COS7 culture supernatant containing soluble IL-5R inhibits the binding of IL-5 to IL-5R expressed on B cells or eosinophils.

Binding Assay of IL-5R to IL-5 (Murine)

The COS7 transformant thus obtained is tested for the production of IL-5R capable of binding to IL-5 using .sup.35 S-methionine and .sup.35 S-cysteine labeled IL-5 (J. Immunol. 140: 1175, 1988; J. Exp. Med. 168: 863, 1988). Binding of the labeled IL-5 is inhibited by the excess amount (100-fold) of the non-labelled IL-5 and thus the cDNA clone pIL-5R.8 is confirmed to code for IL-5R.

Cross-linking of IL-5R to IL-5 and Immunoprecipitation of IL-5R to IL-5 (Murine)

COS7 cells are transfected with pIL-5R.8 followed by cross-linking reaction and immunoprecipitation.

Cross-linking: IL-5R produced by the transformant is tested whether it is the same as those expressed by a IL-5 responsive early B cell, T88-M, by the cross-linking test using .sup.35 S-labeled IL-5 (Proc. Natl. Acad. Sci. USA, 1989, 86: 2311). After electrophoresis, the band pattern on the gel indicates that the molecular weight comparable to IL-5 monomer is decreased (about 22,000) under reduced condition.

Immunoprecipitation: The surface proteins of the transfected cells is .sup.125 I-labeled and immunoprecipitated with anti-IL-5R antibodies, H7 (Int. Immunol. 2: 181, 1990). IL-5R produced by the transformant is found to have a molecular weight of 60,000.

Cell Lines Expressing IL-5R mRNA and the Size of the IL-5R mRNA (Murine)

Poly(A).sup.+ RNA are prepared from IL-5 responsive cell lines such as Y16 (early-B cell), BCL1-B20 (mouse B cell chronic leukemia lymphoma, in vitro line), mouse myeloma cell, MOPC104E, X5568, L cell, IL-3 responsive cell line FDC-P1 derived from mouse bone marrow long-term cultured cell, and IL-2 responsive mouse T cell lines. 2 .mu.g of each of poly(A).sup.+ RNA are tested for the presence of IL-5R mRNA by Northern blot.

Northern blot is carried out using the HindIII-PstI fragment of pIL-5R.8 as a probe (Biochemistry 16:4743, 1977). IL-5 responsive cell lines including Y16, BCL1-B20, MOPC104E are found to express IL-5R mRNA with the size of 5-5.8 kb

Preparation of Poly(A).sup.+ RNA from Human Peripheral Blood Eosinophils (Human)

A DNA sequence coding for human IL-5R is prepared from human peripheral blood eosinophils. Eosinophils are isolated from peripheral blood of healthy volunteers and of a patient with eosinophilia by a density gradient centrifugation using Ficoll (Migita, Y., et al. supra). Whole mRNA is prepared from eosinophils according to the method described by Okayama et al. (ibid). Poly(A).sup.+ RNA is recovered by fractionating the whole RNA with the affinity chromatography using an oligo (dT) cellulose column. One of the poly(A).sup.+ RNA preparation is derived from healthy volunteers and the other is derived from a patient with eosinophilia.

Construction of cDNA Library from mRNA (Human)

The poly(A).sup.+ RNA is reverse transcribed to cDNA using random primers and reverse transcriptase as described above. The cDNA of more than 1.0 kb fragments is selected for cloning. The cDNA fragment derived from eosinophils of healthy volunteers (helv-cDNA) is inserted into the BstXI site of vector pAGS-3 (Miyazaki, et al., 1989, Gene, 79: 269) according to the method described by Seed et al. (ibid). E. coli is then transformed with the recombinant plasmid (helv-cDNA library). The cDNA derived from eosinophils of patients with eosinophilia (eosi-cDNA) is inserted into the EcoRI site of phage .lambda.gt10 using an EcoRI linker. E. coli is then infected with the recombinant phage (eosi-cDNA library).

Screening of Helv-cDNA and eosi-cDNA Libraries for Human IL-5R (Human)

The helv-cDNA library is screened using the HindIII-PstI fragment of pIL-5R.8. A positive clone is isolated and is designated as ph5R. 1. ph5R.1 lacks some of the nucleotide sequence of IL-5R. Subsequently, the eosi-cDNA library is screened using the nucleotide sequence of ph5R.1. Two positive clones designated as HSIL5R and HSIL5R2 are isolated.

Human IL-5R Gene Structure (Human)

FIG. 6 shows restriction maps of the isolated IL-5R cDNAs of HSIL5R and HSIL5R2. The nucleotide sequence was determined according to the Sanger's method (ibid). HSIL5R and HSIL5R2 are membrane bound receptors. The cytoplasmic domain sequence of HSIL5R2 is shorter than that of HSIL5R.

SEQ ID No. 17 and No. 18 show the nucleotiode and deduced amino acid sequence of HSIL5R (420 amino acids in length) and HSIL5R2 (396 amino acid in length), respectively. The amino acid sequence is analyzed as described above.

HSIL5R and HSIL5R2 consist of signal peptide region, extracellular region, transmembrane region, cytoplasmic region. The nucleotide sequence downstream of nucleotide position 1245 distinguishes HSIL5R (SEQ ID No. 17) from HSIL5R2 (SEQ ID No. 18). The amino acid sequence of HSIL5R2 terminates after amino acid Ile (amino acid number 396) located immediately after the nucleotide position 1245.

Expression of Human IL-5R (Human)

Human IL-5R cDNA is inserted into a pCAGGS vector, and COS 7 cells are transfected with the recombinant plasmid. .lambda.gt10 cDNA clones, HSIL5R and HSIL5R2 are digested with EcoRI and the IL-5R cDNA fragment is inserted into the EcoRI site of pCAGGS.

Binding Assay of transfectants with HSIL5R or HSIL5R2 to IL-5 (Human)

The IL-5R expression of the clones are tested using .sup.35 S-methionine- and .sup.35 S-cysteine-labeled murine IL-5 or .sup.125 I-labeled human IL-5. The human IL-5 is prepared as follows:

The IL-5 cDNA fragment is inserted into an expression vector derived from baculovirus. Sf21 cells (Spodotera frugiperda) are infected with the recombinant DNA. The cell are cultured and the culture supernatant is tested for human IL-5 using anti-IL-5 monoclonal antibody, NC17 (Proc. Natl. Acad. Sci. U.S.A. 84: 4581, 1987). The isolated human IL-5 is labeled with .sup.125 I. Binding assay is carried out as described for murine IL-5R

Cross-linking of IL-5R to IL-5 (Human)

IL-5R produced by the positive clones is tested whether it is the same as those produced by eosinophils, by cross-linking test using .sup.35 S-labeled murine IL-5 and .sup.125 I-labeled human IL-5 as described above.

Cell Lines Expressing IL-5R mRNA and the Size of the IL-5R mRNA (Human)

Poly(A).sup.+ RNA may be prepared from human eosinophils, erythroleukemic cell line TF-1, eosinophilic leukemia cell line EoL-3, ATL-2 adult T cell leukemia cell line ATL-2, Burkitt's lymphoma cell line Raji, and histiocytic lymphoma cell line U-937. 6 .mu.g of each of poly(A).sup.+ RNA is tested for the presence of IL-5R mRNA using the entire sequence of HSIL5R cDNA as a probe. Human eosinophils and TF-1cell line are found to express IL-5R mRNA with the size of 1.4 kb and 5.3 kb.

Production of Secretory Human IL-5R

HSIL5R cDNA is inserted into the EcoRI site of Bluescript SK(-). The construct is digested with SalI and KpnI. The SalI-KpnI digested fragment is then incubated with exonuclease III so that the sequence coding for the cytoplasmic domain and transmembrane domain of human IL-5R can be removed. The digested fragment is blunted with mung bean exonuclease followed by a treatment with a klenow fragment and subjected to ligation (Gene 33: 103, 1985). After treatment, a clone is obtained which contains deletion from 3' end to the nucleotide number 995 (SEQ ID No. 17), a site which corresponds to the starting point of deletion in the secretory murine IL-5R cDNA. The deletion mutant is digested with EcoRI and BssHII. The resulting DNA fragment is ligated to a linker containing a stop codon. After ligation, a DNA-linker complex is inserted into an appropriate restriction site of any vector. Alternatively, the HSIL-5R cDNA fragment of the Bluescript SK(-) construct is deleted from 3' end to the nucleotide number 996. As a result of frameshift, the construct contains two stop codons. The secretory human IL-5R construct thus obtained lacks DNA sequences for a cytoplasmic domain and a transmembrane domain and codes for 333 amino acids.

The secretory human IL-5R construct is introduced into host cells and the transfectant produces a secretory human IL-5R. An expression vector is selected according to host cells to be transfected. Host cells include prokaryotes such as gram negative bacteria (E. coli) or gram positive bacteria (Bacillus), yeast, and eukaryotic cell lines derived from insects and mammals.

The following Examples are described for murine secretory IL-5R and membrane type IL-5R.

Preparation of Polyadenylated RNA from Y16 Cell (Murine)

Y16 (2.times.10.sup.7) cells were placed in a 3 liter Spinner culture bottle containing a medium (RPMI 1640, 4% FCS, 5.times.10.sup.-5 M 2-mercaptoethanol,100 U/ml of penicillin, 100 .mu.g/ml of streptomycin) and 300 pg/ml of IL-5. The bottle was sealed and incubated for a week. After incubation, about 5.times.10.sup.9 cells were harvested. 1.times.10.sup.9 cells were solubilized in 50 ml of 5.5M guanidium thiocyanate solution (pH 7.0) containing 25 mM sodium citrate, 0.5% sodium laurylsulcosine, and 0.2M 2-mercaptoethanol according to the method described by Okayama et al. (supra). The cell lysate was layered onto cesium trifluoroacetic acid solution (density: 1.5 g/ml) containing 0.1M EDTA/pH 7.0. The mixture was centrifuged at 15.degree. C., at 125,000 g, for 24 hours. After centrifugation, the RNA pellet was dissolved in distilled water containning 10 mM Tris-HCl/pH 7.5 and 1 mM EDTA. The RNA solution was loaded onto an oligo (dT) cellulose column and the pass-through was loaded onto the column again (Molecular Cloning, 1989, Chapter 7, p 26, Cold Spring Harbor Labolatory Press). The oligo (dT) bounded fraction was eluted and 30 .mu.g of poly(A).sup.+ RNA was recovered.

Construction of cDNA Library in CDM8 (Murine)

30 .mu.g of the poly(A).sup.+ RNA thus obtained was used to synthesize cDNA using a cDNA synthesis kit (BRL, Bethesda, Md.) according to the method described by Seed (supra). The CDM8 vector (see FIG. 2A) was digested with BstXI. After digestion, an approximately 4100 bp fragment was purified by a potassium acetate density gradient centrifugation. The cDNA was ligated to a BstXI linker and a cDNA-linker complex containing cDNA having a size of 1,000 bp or more was selected by a potassium acetate density gradient centrifugation. The fractionated fragments were subjected to ligation with the purified CDM8 vector. E. coli MC1061/P3 was transformed with the construct and about 2 million transformants were obtained as a cDNA library.

Screening of the cDNA Library (Murine)

COS7 (5.times.10.sup.5) cells were placed in each of 100 plates (6 cm). The following day, the COS7 cell was transfected with 2 .mu.g of the plasmid DNA (per plate) prepared from the cDNA library according to the DEAE-dextran method. On day 3, the COS7 cells were removed from the plates and incubated with antibodies, H7 and T21. The COS7 cell was screened for the presence of the H7 and T21 antigens using goat anti-rat IgG antibodies (Panning technique). After screening, plasmid DNAs were prepared from the H7 and T21 antigen positive COS7 cells. Then, E. coli MC106 l/P3 was transformed with the plasmid DNAs. Fresh COS7 cells were fused with the transformants according to the protoplast fusion method. The COS7 cells were screened for the presence of the H7 and T21 antigens according to the Panning technique. After four cycles of the procedure described above, fresh COS7 cells were transformed and the transformant was screened by the Panning technique using goat anti-rat IgG antibody F(ab').sub.2 fragment. This transformation-screening procedure was repeated two times in order to eliminate the contamination of Fc recepter genes. After screening, 50 independent colonies were selected and the plasmid DNA was prepared. Fresh COS7 cells were then transfected with the plasmid DNA and the transformants were tested for the presence of the H7 and T21 antigens. One of the transformants was found to be antigen positive and designated as pIL-5R.8.

The cDNA library prepared from Y16 as described above was screened for the presence of IL-5R cDNA using the fragment inserted in pIL-5R.8 as a probe according to the colony hybridization method (Molecular Cloning, 1989, chapter 1, p90, Cold Spring Harbor Laboratory Press): The HindIII-PstI fragment was prepared from pIL-5R.8 and radiolabeled with .alpha.-.sup.32 P-dCTP according to the random primer method. The transformants of the cDNA library were grown on a solid LB agarose medium (approximately 10,000 colonies per 10 cm plate) overnight. The colonies were transferred to a nitrocellulose membrane and the DNAs on the membrane were hybridized to the radiolabeled prove. Positive colonies were identified through autoradiography. One of the transformants was isolated and designated as pIL-5R.2.

Nucleotide Sequencing of IL-5R (Murine)

The cDNA fragment of pIL-5R.8 was digested with XbaI and inserted into a M13mp19 vector. The construct was digested with BamHI and KpnI. The BamHI-KpnI digested fragment was then digested with exonuclease III: the fragment was digested up to ten minutes with stopping digestion every minute. The digested fragment was blunted with mungbean exonuclease followed by a treatment with a klenow fragment and subjected to ligation (Gene 33:103, 1985). E. coli JM109 was transformed with the constructs to produce different sizes of the M13 deletion mutants. Single-stranded DNAs were prepared from the mutants (Methods in Enzymology 101:58, 1983) and the nucleotide sequence was determined using the M13 primer, 5'-GTTTTCCCAGTCACGAC-3' according to the Sanger's method. Single-stranded DNA was also prepared from the M13 mutant containing the cDNA fragment of a reversed orientation and the nucleotide sequence was determined as described above. The nucleotide sequence thus obtained from the M13 mutant containing the cDNA fragment of a right orientation was found to be complementary to the one of M13 mutant containing the cDNA fragment of a reversed orientation.

SEQ ID No. 15 shows the complete nucleotide and amino acid sequence of pIL-5R.8. The first 17 amino acids are believed to be the signal peptide (Nucleic. Acids. Res. 14:4683, 1986) and the amino acids from amino acid position 340 to amino acid position 361 are considered to be the transmembrane region according to the hydropathy plot. Amino acids 32-34, 128-130,213-215,241-243, 392-394, and 412-414 appear to be sites of N-linked oligosaccharide addition. The previously estimated molecule weight (45,284) of IL-5R differs from the real molecular weight (about 60,000) of IL-5R produced by the COS7 cells transfected with pIL-5.8. The difference of the weight may be due to the addition of N-linked oligosacharide. Nucleotide position 1467 represents the beginning of the deleted nucleotide sequence of pIL-5R.2 The nucleotide sequence of pIL-5R.2 shown in SEQ ID No. 16 was determined using primers(17-mers) synthesized based on the nucleotide sequence of pIL-5 R.8, a T7 primer (5'-ATGGAAATTAATACG-3'), and a primer for the 3' end of CDM8 (5'-TGCAGGCGCAGAACTGG-3') according to the Sanger's method. The pIL-5R.2 is a frame shift mutant, resulting in translation termination to give 4 peptides. The polypeptide encoded by pIL-5R.2 is a secretory IL-5R which is likely to act on B cells or eosinophils in the process of differentiation.

Expression and Binding Test of IL-5R cDNA (Murine)

pIL-5R.8 (CDM8 vector) was digested with XhoI and the IL-5R cDNA fragment was inserted into the XhoI site of pCAGGS vector (see FIG. 2B) whose EcoRI site had been replaced with a XhoI site.

The new construct was designated as pCAGGS.5R. 8. E. coli was transformed with the construct and the transformant was designated as E. coli 5R.8. E. coli 5R.8 was deposited with Fermentation Research Institute, Agency of Industrial Science and Technology and was assigned the accession number FERM BP 3085.

COS7 cells were transfected with pIL-5R.8 or pCAGGS.5R.8 and the cells were harvested two days later. 2-10.times.10.sup.4 cells were incubated with different concentrations of .sup.35 S-labeled IL-5 (2.5.times.10.sup.8 cpm/.mu.g) in the presence or absence of 100-fold excess of non-labeled IL-5 at 37.degree. C. for 10 minutes. After incubation, the number of IL-5 binding per cell was counted and the dissociation constant was calculated.

FIGS. 3(A), (B), (C) shows the Scatchard plot analysis (Ann N.Y. Acad. Sci, 51: 660, 1949) of .sup.35 S-labeled IL-5 binding to IL-5R expressed on transfectants and Y16 cells. The inset shows the direct binding data (.quadrature.: total binding, : non-specific binding). FIG. 3(A) shows the results when COS7 cells were transfected with pIL-5R.8: the dissociation constant was 2 nM and the number of the IL-5 binding was 12,000/cell. FIG. 3(B) shows the results when COS7 cells were transfected with pCAGGS.5R.8: the dissociation constant was 9.6 nM and the number of the IL-5 binding was 880,000/cell. FIG. 3(C) shows the results when Y16 cells were also tested for the IL-5 binding. A high affinity IL-5R and a low affinity IL-5R were found in the Y16 cells. The high affinity IL-5R has the number of IL-5 binding of 1,200/cell with the dissociation constant (KD) of 20 pM. The low affinity IL-5R has the number of IL-5 binding of 22,000/cell with the dissociation constant (KD) of 5.1 nM. These results suggest that the inserted IL-5R cDNA encodes a low affinity IL-5 recepter.

Cross-linking Test of Low Affinity IL-5R to IL-5 (Murine)

Because COS7 cells transfected with pCAGGS.5R.8 were found to express IL-5R at a higher level than those transfected with pIL-5R.8, pCAGGS.5R.8 was used for the following experiments.

COS7 cells (1.times.10.sup.5) were transfected with pCAGGS or pCAGGS5R.8 and the transformants were incubated with 4 nM .sup.35 S-labeled IL-5 in the presence or absence of 100-fold excess of non-labeled IL-5 as follows: a pCAGGS transformant without non-labeled IL-5 (lane 1), a pCAGGS transformant with non-labeled IL-5 (lane 2), a pCAGGSSR.8 transformant without non-labeled IL-5 (lane 3, 5), a pCAGGS5R.8 transformant with non-labeled IL-5 (lane 4, 6). The mixture was incubated at 37.degree. C. for 10 minutes. Cells were washed extensively and then disuccinimidyl tartarate (DST)(Piece Chemical, Rockford, Ill.) was added to the cell suspension. The cell suspension was incubated at 4.degree. C. for 30 minutes and then 1% Triton X-100 was added to the suspension to disrupt the cells. The disrupted cell suspension was loaded on a 7.5% SDS-polyacrylamide gel in the reducing (lane 5, 6) or non-reducing (lane 1-4) conditions.

After electrophoresis, the gel was analyzed with Bio-Analyzer 100 (Fuji Film). The results are shown in FIG. 4. A band of approximately 90-100 KD in size was found which could be a low affinity IL-5R previously reported by Mita, et al., in Proc. Natl. Acad. Sci. U.S.A. 86: 2311, 1989. In contrast, the molecular weight of the band in the reducing condition was about 75 KD (lane 5 in FIG. 4). The difference was due to the dissociation of monomeric .sup.35 S-labeled IL-5 (MW: 22,000) from the IL-5-IL-5R complex, because biologically active IL-5 binds to its receptor as a disulfide-linked dimer.

Immunoprecipitation of IL-5R expressed on pCAGGS.5R.8 Transfected COS7 (Murine)

The surfaces of the pCAGGS.5R.8 transfected COS7 (5.times.10.sup.6) cells were labeled with .sup.125 I using Iodobeads (Pierce Chemical, Rockford, Ill.). The cell was disrupted and H7 antibody was added to the cell lysate. Protein G-Sepharose (Pharmacia, Piscataway, N.J.) was added to the mixture and the mixture was incubated at 4.degree. C. for 12 hours. The proteins adsorbed on the Sepharose was loaded on the SDS-PAGE. After electrophoresis under a reducing condition, the gel was analyzed with Bio-Analyzer 100. The band (MW: about 60 KD) was found only in the lane where the sample was prepared from the cell transfected with pCAGGS.5R.8(FIG. 5).

Purification and Amino acid Sequence Analysis of Secretory IL-5R (Murine)

The IL-5R cDNA fragment obtained by XhoI digestion of pIL-5R.2 was inserted into pCAGGS vector by the similar method as in the case of pIL-5R.8 and the construct was designated as pCAGGS.5R.2. E. coli was transformed with the pCAGGS.5R.2 and the transformant was designated as E. coli 5R.2. E. coli 5R.2 was deposited with Fermentation Research Institute, Agency of Industrial Science and Technology and was assigned the accession number FERM BP 3084.

COS7 cells were transfected with pCAGGS.5R.2 DNA according to the DEAE dextran method and was cultured in serum free medium (Iscove's DMEM) for two days. The culture supernatant was concentrated and the concentrate was electrophoresed on SDS-PAGE. A band (MW: approximately 50,000) was found in the lane on which the culture supernatant of pCAGGS.5R.2 transfected COS7 was loaded, while no band was found in the lane on which the culture supernatant of the pCAGGS vector alone was loaded. The culture supernatant of the pCAGGS.5R.2 transfected COS7 was loaded onto a column filled with H7 anibodies bound glycosylhard-gel (Seikagaku Kogyo, Tokyo). The column was washed with 2 mM HEPES solution containing 0.1% CHAPS and then H7 bound fractions were eluted out with 350 mM acetic acid. The fractions were lyophilized and then solubilized in a sample buffer for SDS-PAGE. The mixture was electrophoresed according to the method described by Laemmli in Nature 227: 680, 1970. The protein on the gel was transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, Mass.) according to the electroblotting method. The band corresponding to a molecular weight of about 50,000 was cut out of the membrane and analyzed with a gas phase sequencer 447A (with HPLC system, Applied Biosystem Co.). The amino acid sequence of the N terminus of secretory IL-5R was as follows: Asp-Leu-Leu-Asn-His-Lys-Lys-Phe-Leu-Leu-Leu-Pro-Pro-Val-X-Phe-Thr-Ile-Lys- Ala. This amino acid sequence was found to be the same one (amino acid number 18-37) deduced from the nucleotide sequence of pIL-5R.8, membrane bound IL-5R cDNA clone. The amino acid sequence (amino acid number 1-17) is believed to be a signal peptide. X (amino acid number 15) may be Asn, which is deduced from the nucleotide sequence of cDNA, and to which a N-linked oligosaccharide is believed to bind.

The following Examples are described for human IL-5R.

Preparation of human Poly(A).sup.+ RNA

Eosinophils were obtained from 28 liter of peripheral blood of healthy volunteers and 50 ml of peripheral blood of a patient with eosinophilia. After removing erythrocytes, fractions containing eosinophils (1.09 g/ml) were collected from each sample by a density gradient centrifugation using Ficoll. The fraction contained 50% eosinophils and the number of eosinophis was 2.8.times.10.sup.9 [healthy volunteers (helv)] and 2.0.times.10.sup.9 [eosinophilia (eosi)]. 5 .mu.g of poly(A).sup.+ RNA was recovered from each cell source as described above.

Construction of Human IL-5R cDNA Library

5 .mu.g of each poly(A).sup.+ RNA thus obtained was used to synthesize cDNA (helv-cDNA, eosi-cDNA) as described above. The helv-cDNA was ligated to a BstXI linker and a fragment of helv-cDNA-linker complex having a size of 1,000 bp or more was selected. The fragment was then inserted into a BstXI digested pAGS-3 vector. E. coli MC1061 was transformed with the recombinant plasmid and about one million transformants were obtained (helv-cDNA library). The eosi-cDNA was ligated to a EcoRI linker and fragments of eosi-cDNA-linker complex having a size of 1,000 bp or more were selected. The fragments were inserted into a EcoRI digested .lambda.gt10 vector. E. coli C600Hfl was infected with the recombinant phage and 1.6 million independent plaques were obtained (eosi-cDNA library).

Screening of helv- and eosi-cDNA libraries according to the colony-hybridization method

The helv-cDNA library was screened according to the colony hybridization method. One million colonies of the helv-cDNA library were grown on a solid medium and the colonies were transferred to 100 sheets of nitrocellulose membranes (8 cm in diameter). After DNA fixation, the membrane was placed in a bag containing 10.times. Denhardt's solution, 6.times. SSC (0.9M NaCl, 0.09M sodium citrate), 100 .mu.g/ml of heat-denatured salmon sperm DNA. The .sup.32 P-labeled, 1.2 kb HindIII-PstI fragment of pIL-5R.8 was added to the bag and hybridization was carried out at 65.degree. C. for 24 hours under less stringent conditions. The membrane was washed at 45.degree. C. in a solution containing 1.times. X SSC and 0.1% SDS. After washing, a X-ray film was overlayed on the membrane for autoradiography as described above. A positive clone was obtained and was designated as ph5R.1. However, the cDNA fragment of ph5R.1 was found to contain only 1.0 kb, which was not an right size for IL-5 R. Subsequently, the eosi-cDNA library was screened using the XhoI digested, 1.0 kb fragment of ph5R.1 as a probe according to the protocol of Colony/Plaque Screen. Approximately one million clones of the eosi-cDNA library was grown on a solid medium and the plaques were transferred to nylon membranes (13 cm in diameter, Colony/Plaque Screen, Dupont-NEN, Boston, Mass.). Hybridization was carried out at 65.degree. C. for 24 hours in a solution containing 1% SDS, 1M NaCl, 10% Dextran sulfate, 100 .mu.g/ml of heat-denatured salmon sperm DNA. After hybridization, the membrane was washed at 65.degree. C. for an hour in a solution containing 2.times. X SSC and 1% SDS. Two positive clones containing about 2 kb cDNA fragment were obtained and designated as HSIL5R and HSIL5R2.

E. coli was transformed with HSIL5R or HSIL5R2 and the transformants were designated as E. coli HSIL5R or E. coli HSIL5R2, respectively.

The transformants were deposited with Fermentation Research Institute, Agency of Industrial Science and Technology and were assigned the accession number as follows:

DNA Sequence Analysis of HILS5R and HSIL5R2

HILS5R and HSIL5R2 were digested with EcoRI, and the EcoRI digested IL-5R fragment was inserted into the EcoRI site of Bluescript KS (-) vector (Stratagene, La Jolla, Calif.). The nucleotide sequence was determined according to the Sanger's method. The sequence was determined in both 5' and 3' direction. Initial primers were synthesized according to the sequence of the 5' upstream of the IL-5R cDNA fragment (T3 primer) and of the 3' downstream of the IL-5R cDNA fragment (T7 primer). After the 5' and 3' end sequences were determined, subsequent primers were synthesized according the sequence analyzed by the DNA sequencing. The nucleotide sequence thus determined was found to be complementary.

SEQ ID No. 17 shows the nucleotide and the corresponding amino acid sequence of HSIL5R. The first, 20 amino acids are hypothetically a signal peptide and amino acids 345 to 365 are believed to be a transmembrane region according to hydropathy plot. These assumption are based on the same model as those of mouse. Amino acids 35-37, 131-133, 137-139, 142-144, 216-218, and 244-246 seem to be the sites of N-linked oligosaccharide addition. The estimated molecular weight (45,556) of IL-5R from cDNA clone differs from the real molecular weight (about 60,000) of IL-5R produced by the transformed COS7 cell. The differenece of the weight may be due to the N-linked oligosaccharide. The nucleotide sequence downstream of nucleotide position 1245 distinguishes HSIL5R (SEQ ID No. 17) from HSIL5R2(SEQ ID No. 18).

SEQ ID No. 18 shows the nucleotide and the corresponding amino acid sequence of HSIL5R2. The amino acid sequence of HSIL5R2 terminates at Ile (amino acid number 396), while HSIL5R contains additional 24 amino acids following Ser at amino acid No. 396. The amino acid sequences of HSIL5R and HSIL5R2 are identical from Met (amino acid number 1 ) to Gly (amino acid number 395) except for an amino acid at position 129 where the amino acid is Val on the sequence of HSIL5 and Ile on the sequence of HSIL5R2.

Expression of Human IL-5R on COS7 and Cross-linking experiment

The Bluescript kS (-) recombinant was digested with EcoRI. The restriction fragments containing IL-5R of HSIL5R and HSIL5R2 were inserted into pCAGGS. The resulting constructs were designated as pCAGGS.HSIL-5R and pCAGGS.HSIL5R2. COS7 cells were transfected with these recombinant DNAs and the transformed cells were tested for their chemical characteristics using .sup.35 S-labeled murine IL-5 or .sup.125 I-labeled human IL-5 (2.times.10.sup.6 cpm/.mu.g) according to the cross-linking method.

Binding of .sup.125 I-labeled human IL-5 to IL-5R expressed on the COS7 cell (pCAGGS.HSIL5R transformant) was shown in the inset of FIG. 7A, and the results analyzed by Scatchard plot was shown in FIG. 7A. Binding of .sup.125 I-labeled human IL-5 to IL-5R expressed on COS7 cell (pCAGGS.HSIL5R2 transformant) was shown in the inset of FIG. 7B, and the results analyzed by Scatchard plot was shown in FIG. 7B. Binding of .sup.35 S-labeled murine IL-5 to IL-5R expressed on the COS7 cells (pCAGGS.HSIL5R transformant) was shown in the inset of FIG. 7C, and the results analyzed by Scatchard plot was shown in FIG. 7C. Binding of .sup.35 S-labeled murine IL-5 to IL-5R expressed on COS7 cell (pCAGGS.HSIL5R2 transformant) was shown in the inset of FIG. 7D, and the results analyzed by Scatchard plot were shown in FIG. 7D.

A high affinity IL-5R with the dissociation constant (KD) of less than 100 pM was not detectable by .sup.125 I-labeled human IL-5 because of poor specific radioactivity. To calculate the dissociation constant of a high affinity IL-5R, we used .sup.35 S-labeled mouse IL-5 which has high specific radioactivity and is not denatured. The dissociation constant of the pCAGGS.HSIL5R transfected COS7 cells were about 590 pM when radiolabeled human IL-5 was used, while the dissociation constant of the same pCAGGS.HSIL5R transfected COS7 cells were about 250 pM when radiolabeled mouse IL-5 was used. The dissociation constant of the pCAGGS.HSIL5R2 transfected COS7 cells were about 410 pM with radiolabeled human IL-5, while the dissociation constant of the same pCAGGS.HSIL5R2 transfected COS7 cells were about 355 pM when radiolabeled mouse IL-5 was used. These results are comparable to the dissociation constant (170-330 pM) of eosinophils from healthy adult peripheral blood that we reported previously. The data of the previous report were calculated by Scatchard analysis of binding assays using .sup.35 S-labeled mouse IL-5.

The dissociation constant thus determined was higher than that of mouse low affinity IL-5R and fell into the average value of normal human eosinophils. Taken altogether, the isolated IL-5R cDNA fragment was expressed on the surface of the COS7 cells and the IL-5R expressed on the cell surface are responsible for the binding of human IL-5.

FIG. 8 shows inhibitory effects of cytokines on the binding of IL-5 to IL-5R. IL-5R expressed on the COS7 transformants specifically binds to human and mouse IL-5 but not to human IL-2, human IL-3, human IL-4, human IL-6, human GM-CSF or human G-CSF.

Cross-linking of Radiolabeled IL-5 to the COS7 Transformants

The COS7 transformant (1.times.10.sup.5 cells) carrying pCAGGS.HSIL5R or pCAGGS.HSIL5R2 and either 5.5 nM .sup.35 S-labeled mouse IL-5 or 1 nM .sup.125 I-labeled human IL-5 were mixed in the presence or absence of 250-fold excess of non-labeled IL-5. After one hour incubation at 4.degree. C., 1 mM bis(sulfosuccinimidyl) suberate (Pierce Chemical Co., Rockford, Ill.) was added to the mixture. The mixture was further incubated at 4.degree. C. for 30 minutes. After the incubation, binding was analyzed as described above.

In FIG. 9, COS7 cells transfected with a pCAGGS vector alone or pCAGGSHSIL5R were incubated with .sup.35 S-labeled murine IL-5 (A; lane 1, 2, 3) or .sup.125 I-labeled human IL-5(B; lane 4, 5, 6). COS7 cells transfected with a pCAGGS.HSIL5R were incubated with .sup.35 S-labeled murine IL-5 in the presence (lane 3) or absence (lane 2) of 250-fold excess of non-labeled IL-5, or were incubated with .sup.125 I-labeled human IL-5 in the presence (lane 6) or absence (lane 5) of 250-fold excess amount of non-labeled IL-5.

There were two bands corresponding to about 105 kD (lane 2) and 86 kD (lane 5). Since murine IL-5 is 45 kD and human IL-5 is 31 kD, the molecular weight of human IL-5R could be estimated to be 55,000-60,000. This molecular weight of human IL-5R is almost the same as that of IL-5R expressed on eosinophils as we reported previously (Cellular Immunology, 133; 484-469). In the presence of a 250-fold excess of non-labeled IL-5, no band was found (lanes 3, 6 in FIG. 9).

Same experiment was carried out using pCAGGS.HSIL5R2 and the results were very similar to that described above.