NO323195B1 - Protein having TPO activity, DNA coding for said protein, production methods, host cells, pharmaceutical compositions, uses and antibody - Google Patents

Description

The present invention relates to new proteins that specifically stimulate or

improves proliferation and differentiation of megakaryocyte precursor cells. The invention further relates to DNA sequences that code for said proteins, methods for producing said proteins, and host cells that have been transformed or transfected with said DNA sequence. Pharmaceutical compositions comprising said protein are also part of the present invention. Furthermore, the invention also includes the use of said polypeptide and antibody that specifically binds said proteins.

Megakaryocytes are large cytoplasm-rich many-nucleated cells that produce blood-

plates and which are mainly found in bone marrow. Megakaryocytes originate from a pluripotent hematopoietic stem cell in the bone marrow. A primitive pluripotent stem cell differentiates to some extent into a megakarocyte progenitor cell, which is committed to the megakarocyte lineage. The megakaryocyte precursor cells proliferate and differentiate into megakaryocytes. Megakaryocytes undergo further polyploidization and cytoplasmic maturation and finally anuclear cytoplasmic fragments, namely platelets, are released into the circulatory system. On average, 2 to 4,000 platelets are formed from a mature megakaryocyte.

Although the mechanism of platelet formation is still unclear in many respects, it is recognized that the megakaryocytes are typically located on the albuminal surface of the bone marrow sinus endothelium and produce cytoplasmic processes that extend into the sinusoid where they undergo platelet fragmentation.

There are many unclear points regarding the mechanism of generating platelets although it is likely that the megakaryocytes are locally present in the dermal layer of the skin of the venous sinus in the bone marrow and that the cytoplasm passes through the dermal skin resulting in a thread-like projection on the inner wall of the venous sinus whereby the platelets are emptied.

It has been suggested that there is a specific function for the production and control of mega-karocyte hematopoiesis platelet production. In healthy humans and animals, the number of effective platelets is maintained although it is known that, for example, by administrating

ing of an anti-platelet antibody to healthy animals, the number of platelets decreases rapidly and they begin to increase temporarily more than usual, eventually returning to normal levels. Also within the clinical field, it has been known that a decrease in the number of platelets thrombocytopenia and thrombocytosis even when the number of erythrocytes and leukocytes is at a normal level.

The most important function of a platelet is the production of a thrombus in a haemostatic mechanism. If the haemostatic mechanism does not function properly due to a decrease in the number of platelets, this results in a tendency towards haemostasis. The presence of a specific regulatory mechanism has been suggested with respect to megakaryocytopoiesis and thrombopoiesis. Platelets are maintained in effective numbers in healthy humans and normal animals. However, it is known that when an anti-platelet antibody was administered to a normal animal, only the platelet count rapidly decreases for a short period of time, and then begins to increase and temporarily exceeds the normal level, but eventually returns to that normal level. It is also known in the clinical field that a decrease in the number of platelets (thrombocytopenia) or an increase in the number (thrombocytosis) occurs even when the number of erythrocytes and leukocytes is normal. To date, however, no success has been reported in isolating and identifying a specific regulatory factor involved in platelet production (for example, with similarity in the case of erythropoietin in erythrocyte formation).

The most important function of platelets is the formation of a blood clot in the hemostatic mechanism. Bleeding tendencies occur when the normal function of the haemostatic mechanism is impaired by thrombocytopenia. In the field of radiotherapy and chemotherapy of cancer, thrombocytopenia caused by bone marrow suppression is a fatal complication, and platelet transfusion is used in such patients to prevent bleeding tendency. Platelet transfusion was also used for patients after bone marrow transplantation or from aplastic anaemia.

Platelets for use in such platelet transfusion are prepared by platelet pheresis from blood in healthy blood donors, but such platelets for transfusion use have a short shelf life and a possibility of causing contamination by bacterial infection. Platelet transfusion also has a possible risk of exposing patients to serious viruses such as human immunodeficiency virus (HIV) or various hepatitis viruses, including antibodies specific for a major histocompatibility antigen (HLA) or transfused platelets or causing graft versus host disease (GVHD) due to contaminated lymphocytes -cytes in the platelets for transfusion use.

As a consequence, it would be of great advantage if the internal platelet formation could be stimulated in thrombocytopenic patients, and at the same time their platelet transfusion tendency could be reduced. If, in addition, thrombocytopenia in cancer patients undergoing radiotherapy or chemotherapy can be corrected or prevented, such treatments can be made safer, the intensity of the treatment can probably be increased, and further improvement of anti-cancer effects can be expected.

For these reasons, a large number of intensive studies have been carried out on the isolation and identification of specific regulatory factors involved in the regulation of megakaryocyte and platelet production. According to in vitro studies, regulatory factors that control megakaryocytopoiesis are roughly divided into the following two factors (cf. Williams et al., J. Cell. PhysioL, vol. 110, pp. 101-104 (1982). Megakaryocyte colony stimulating factor (Meg -CSF) is a regulatory factor that stimulates proliferation and differentiation of CFU-MK to form megakaryocyte colonies in a semi-solid culture medium. The other regulatory factor, cold megakaryocyte potentiating factor (Meg-Pot), megakaryocyte stimulating factor, thrombopoiesis stimulating factor or similar acts mainly on immature or mature megakaryocytes and thus improves differentiation and maturation. Meg-Pot is detectable in combination with Meg-CSF activity in some cases. Since, in addition, platelet counts increase when serum or plasma collected from experimentally induced thrombocytopenic animals is administered to other normal animals, it has been suggested that it is a humoral factor, called thrombopoietin (TPO) that has the ability to promote platelet production tion in vivo.

In recent years, some cytokines whose genes have been cloned have been investigated for their ability to stimulate megakaryocytopoiesis and thrombopoiesis. Human IL-3 stimulated human megakaryocyte colony formation (Bruno et al., Exp. Hematol., vol. 16, pp. 371-377, 1988) and, at least in monkeys, increased platelet count (Donahue et al., Science vol. 241, pp. 1820, 1988). However, since IL-3 is a factor that exerts its effects on proliferation. and differentiation of all hematopoietic cells, it can be dissociated from specific regulatory factors that control megakaryocytopoiesis and platelet production. Human IL-6 did not show Meg-CSF activity, but acted on immature megakaryocytes and thereby enhanced their differentiation into mature megakaryocytes (Williams et al., Exp. Hematol., vol. 18, p. 69, 1990). In vivo administration of IL-6 induced platelet production and promoted both maturation and the shift to higher ploidy of bone marrow megakaryocytes in primates, but also caused side effects such as reduction of body weight, induction of acute phase protein (Asano et al., Blood, vol. 75, p 1602-1605, 1990; Stahl et al., Blood, vol. 78, pp. 1467-1475, 1991). Human BL-11 showed no Meg-SCF activity, but exhibited Meg-Pot activity and promoted platelet production in mice (Neben et al., Blood, vol. 81, pp. 901-908, 1993). In addition, human ILF significantly increased platelet counts in primates (Mayer et al., Blood, vol. 81, pp. 3226-3233, 1993), but its in vitro effect on megakaryocytes was weak (Burstein et al., J. Cell. PhysioL , vol. 153, pp. 305-312, 1992).

Although clinical application of these cytokines as platelet-enhancing factors is expected, their function is not specific for megakaryocyte binding and they cause side effects. Development of a platelet-increasing factor which is specific for the megakaryocyte-platelet system and which ensures few side effects has been desired within the clinical field. Meg-SCF, Meg-pot or TPO activity has been found in the serum, plasma or urine of thrombocytopenic patients or animals, or in the culture supernatant of certain human cultured cell lines. Whether these activities are due to the presence of a single factor or a combination of several factors or whether they differ from known cytokines is currently unknown.

Hoffman et al, have found that sera from patients with aplastic anemia and amegakaryocytic thrombocytopenia purpura contained a Meg-SCF activity that significantly increased human megakaryocyte colony formation (Hoffman et al., N. Eng. J. Med., vol. 305, p .553-538, 1981). Subsequently, Mazur et al., reported that Meg-CSF activity present in the serum of aplastic anemia patients was different from IL-3 and GM-CSF (Mazur et al., Blood, vol. 76, pp. 290-297, 1990 ). Similar Meg-CSF activity has been found in sera from cancer patients receiving intensive cytotoxic chemotherapy and bone marrow transplant patients (Mazur et al., Exp. Hematol., vol. 12, pp. 624-628, 1984; de Alarcon and Schmeider , Prog. Clin. BIo. Res., vol. 215. pp. 335-340, 1986). Hoffman et al., have reported that Meg-CSF with an apparent molecular weight of 46,000 was purified from the plasma of patients with hypomegakaryocytic thrombocytopenia (Hoffman et al., J. Clin. Invest., vol. 75, pp. 1174-1182, 1985 ), but it was found upon further study that the material was not at a level of purity that would allow accurate amino acid sequencing (Hoffman, Blood, vol. 74, pp. 1196-1212, 1989). A substance that has TPO-like activity that enhances the incorporation of <75>se-selenomethionine into newly formed platelets in mice has been specifically purified from the plasma of thrombocytopenia patients or the urine of patients with iodopathic thrombocytopenia purpura (ITP), and the apparent molecular weight of plasma-derived factor was determined to be 40,000 (Grossi et al., Hematologica, vol. 72, pp. 291-295, 1987; Vannucchi et al., Leukemia, vol. 2, pp. 236-240, 1988).

Meg-CSF and TPO-like activities have also been demonstrated in urine samples from patients with aplastic anemia and severe ITP (Kawakita et al., Br. J. Haematol., vol. 48, pp. 609-615, 1981; Kawakita et al. ., Blood, vol. 556-560, 1983). Kawakita et al. have further reported that Meg-CSF activity found in urine extract from aplastic anemia patients showed an apparent molecular weight of 45,000 by gel filtration under dissociation conditions (Kawakita et al., Br. J. Haematol., vol. 62, pp. 715-722, 1986). Erikson-Miller et al., have also reported on the purification of Meg-SCF from a corresponding urine sample, and no information on structure (Erikson-Miller et al., "Blood Cell Growth Factors: their present and future use in hematology and oncology" ed .by Murphy, AlphaMed Press, Dayton, Ohio, pp. 204-220, 1992). Turner et al. have purified a megakaryocyte-stimulating factor (MSF) having Meg-SCF activity from urine of bone marrow transplant patients and cloned the gene (Turner et al., Blood, vol. 78, p. 1106 279a, (abstr., suppl. 1)). This MSF has a molecular weight from 28,000 to 35,000. The identity of this factor with Meg-CSF detected in serum and plasma samples of thrombocytopenic patients and its platelet-enhancing activity remains to be deduced.

A substance having a TPO-like activity with a molecular weight of 32,000 has been purified from the culture supernatant of a human embryonic kidney-derived cell line (HEK cells) and its biological and biochemical properties have been under extensive investigation, but its structure is still unknown (McDOnald et al., J. Lab. Clin. Med., vol. 106, pp. 162-174, 1985; McDonald, Int. J. Cell. Cloning, vol. 7, pp. 139-155, 1989) . Other researchers have reported that the main activity in conditioned medium of HEK cells, which enhances megakaryocyte maturation in vitro, is due to known cytokines, namely IL-6 and EPO (Withy et al., J. Cell. PhysioL vol. 15, pp. 362-372 , 1992).

With respect to factors derived from animals, Evatt et al. reported that a TPO-like activity that enhances incorporation of ^Se-selenomethionine into newly formed platelets in rabbits and mice was found in the plasma of thrombocytopenic rabbits induced by antiplatelet serum injection (Evatt et al., J. Lab. Clin. Med., vol .83, pp. 364-371, 1974). In addition to this, a number of similar studies have been reported from 1960 to 1970 (for example, Odell et al., Proe. Soc. Biol. Med., vol. 108, pp. 428-431, 1961; Evatt and Levin, J. Clin. Invest., vol. 48, pp. 1615-1626, 1969; Harker, Am. J. PhysioL vol. 218, pp. 1376-1380, 1970; Shreiner and Levin, J. Clin. Invest., vol. 49 pp. 1709-1713, 1970; and Penington, Br. Med. J., vol. 1, pp. 606-608, 1070). Evatt et al., and Hill and Levin have carried out partial purification of a TPO-like activity from the plasma of thrombocytopenic rabbits (Evatt et al., Blood, vol, 54, pp. 377-388, 1979; Hill and Levin, Exp. Hematol., vol. 14, pp. 752-759, 1986). Purification of this factor was then continued, and recording of an activity to improve the differentiation of maturation of the megakaryocytes in vitro, namely the Meg-Pot activity, to show that this activity has an apparent molecular weight of 40,000 to 46,000 when determined by gel filtration (Keller et al., Exp. Hematol., vol. 16, pp. 262-267, 1988; Hill et al., Exp. Hematol., vol. 20m pp. 354-360, 1992). Since IL-6 activity was not detectable in the plasma of rabbits with severe acute thrombocytopenia induced by antiplatelet serum administration, it was argued that this TPO-like activity is due to a factor different from IL-6 (Hill et al., Blood, vol. 80 , pp. 346-351, 1992).

Tayrien and Rosenberg have also purified a factor having a molecular weight of 15,000 from thrombocytopenic rabbit plasma and culture supernatant from HEK cells, which stimulates the production of platelet factor 4 in a rat megakaryocytic cell line, but no information regarding its structure (Tayrien and Rosenberg, J. Biol. Chem., vol. 262, pp. 3262-3268, 1987).

In addition, Nakeff has run a Meg-CSF activity in the serum of thrombocytopenic mice induced by anti-platelet administration (Nakeff, "Experimental Hematology Today" ed. by Baum and Ledney, Springer-Verlag, NY pp, 111-123, 1977). On the other hand, sera from thrombocytopenic rabbits enhanced maturation of megakaryocytes (Keller et al., Exp. Hematol., vol. 16, pp. 262-267, 1988; Hill et al., Exp. Hematol., vol. 19, p . 903-907, 1989) and stimulated morphological change of megakaryocytes to platelets (Leven and Yee, Blood, vol. 69, pp. 1046-1052, 1989), but had no detectable Meg-CSF activity.

Miura et al. have demonstrated Meg-CSF activity in plasma from rats rendered thrombocytopenic by sublethal whole-body irradiation (Miura et al., Blood, vol. 63, pp. 1060-1066, 1984), and claim that induction of Meg-CSF activity in vivo is related to a decrease in megakaryocytes, but not a decrease in platelets, because this activity is not altered by platelet transfusion (Miura et al., Exp. Hematol., vol. 16, pp. 139-144, 1988). Mazur and South have demonstrated a Meg-CSF activity in the serum of sublethally irradiated dogs and reported that this factor has an apparent molecular weight of 175,000 when measured by gel filtration (Mazur and South, Exp. Hematol., vol. 13, p .1164-1172,1985). In addition to the above, serum-, plasma- and urine-derived factors have been reported by other researchers, for example by Straneva et al. (Straneva et al., Exp. Hematol., vol. 15, pp. 657-663, 1987).

Thus, as described above, various activities that stimulate megakaryocytopoiesis and thrombopoiesis have been found in biological samples derived from thrombocytopenic patients and animals, but isolation of these factors, their biochemical and biological identification and determination of their characteristic features have not been achieved due to the extremely small amounts in natural sources such as blood and urine.

The objectives of the present invention are to isolate a TPO protein from natural sources and identify it, the TPO protein that has an activity that stimulates or increases platelet production in vivo and/or to improve proliferation and differentiation of megakaryocyte progenitor cells (hereinafter referred to as "TPO activity "); to isolate a gene that codes for the TPO protein and provide a method for the production of said protein in a homogeneous quality and in a large quantity with recombinant DNA techniques. Success in achieving such goals will lead to the replacement or frequency reduction of the currently used platelet transfusion, and such a new protein will also be used for the treatment and diagnosis of platelet disorders.

The present invention therefore concerns (i) a purified and isolated DNA sequence that codes for a protein that has TPO activity, which is selected from the group consisting of:

(a) DNA sequences shown in SEQ JD NO: 194,195 and 196, or complementary strands thereof; and (b) DNA sequences that hybridize under stringent conditions to the DNA sequences as defined in (a) or fragments thereof; and (c) DNA sequences that will hybridize to the DNA sequence defined in (a) and (b) but for the degeneracy of the genetic code, (ii) a method for the production of a protein having TPO activity, and comprising the step of: culturing under suitable nutritional conditions, prokaryotic or eukaryotic host cells transformed or transfected with said DNA sequence in a manner that enables expression of the protein; and isolating the desired protein product obtained by expression of said DNA sequence, and (iii) a protein product obtained by expression of a prokaryotic or eukaryotic host cell of said DNA sequence.

The present invention further relates to a pharmaceutical composition which comprises an effective amount of said protein which has TPO activity, and a method for treating platelet disorders, particularly thrombocytopenia, comprises administration of said protein to patients who have the disorders. Figure 1 shows sephacryl S-200HR gel filtration chromatography of phenyl sepharose 6 FF/LS F2 derived from XRP. Figure 2 shows Capcell Pak Cl reverse phase chromatography of YMC-pack CN-AP TPO-active fraction derived from low molecular weight TPO sample (Sephacryl S-200HR F3) of XRP. Figure 3 shows an SDS-PAGE analysis of the Capcell Pak Cl TPO-active fraction (FA) from the low molecular weight TPO sample of XRP. Figure 4 shows the peptide map on C18 reverse phase HPLC of rat TPO isolated by SDS-PAGE. The peptide fragment was obtained by systematic hydrolysis with three proteases.

Figure 5 shows TPO activity derived from XRP in the rat CFU-MK assay system.

Figure 6 shows a construction of expression vector pEF18S.

Figure 7 shows TPO activity in culture supernatant of COSI cells in which pEF218S-A26 was introduced into the rat CFU-MK analysis system. Figure 8 shows TPO activity in culture supernatant of COSI cells in which pEF18S-HL34 was introduced into the rat CFU-MK analysis system. Figure 9 shows TPO activity in culture supernatant of COSI cells in which pHTl-231 was introduced into the rat CFU-MK analysis system. Figure 10a shows TPO activity in culture supernatant of COSI cells in which pHTF1 was introduced into the rat CFU-MK analysis system. Figure 10b shows TPO activity in the culture supernatant of COSI cells in which pHTF1 was introduced into the rat M-07e analysis system. Figure 11 shows a restriction map of OHGT1 clone and construction of pHGT1 and FHGTE.E:EcoRL H:HindIIL S:Sall) Figure 12a shows TPO activity in culture supernatant of COSI cells in which pEFHGTE was introduced into rat CFU-MK analysis system. Figure 12b shows TPO activity in culture supernatant of COSI cells in which pEFHGTE was introduced into the M-07e analysis system. Figure 13a shows TPO activity in culture supernatant of COSI cells in which pHTl-211#1, pHTl-191#1 or pHTl-171#2 were introduced into the rat CFU-MK analysis system. Figure 13b shows TPO activity in the culture supernatant of COSI cells in which pHTl-163#2 was introduced into the rat CFU-MK analysis system. Figure 14 shows TPO activity in culture supernatant of COS 1 cells in which pHTl-211#1, pHTl-191#1, pHTl-171#2, or pHTl-163#2 were introduced into the M-07e analysis system. Figure 15 is a chromatogram of a reverse phase chromatography (Vydac C4 column) for the purification of human TPO from culture supernatant of CHO cells into which pDEF202-hTPO-P1 has been introduced and allowed TPO to be expressed. Figure 16 is a photograph showing SDS-PAGE separation of human TPO purified from the culture supernatant of CHO cells into which pDEF202-hTPO-P1 has been introduced and allowed TPO to be expressed. Figure 17 is a chromatogram of a reverse phase chromatography for purification of human TPO from E. coli into which pCFM536/h6T(1-163) has been introduced and allows TPO to be expressed. Figure 18 is a photograph showing SDS-PAGE separation of variant human TPO, h6T(1-163) isolated and purified from E. coli into which pCFM536/h6T(1-163) has been introduced and allowed TPO to be expressed. Figure 19 shows an elution pattern of hTP0163 on Superdex 75 pg column with regard to purification of hTPO 163 from the culture supernatant, as a starting material, prepared by transfection of human TPO expression plasmid pDER202-hTPO 163 into CHO cells. The amount of protein was determined at 220 nm. Figure 20 shows an SDS-PAGE analysis of standard hTP0163 eluted on Superdex 75 pg column by purification of hTPO 163 from the culture supernatant, as a starting material, prepared by transfection of human TPO expression plasmid pDEF202-hTPO 163 in CHO cells. hTPO 163 was silver stained on the gel.

Figure 21 shows a structure of the expression vector pSMT201.

Figure 22 is a graph showing TPO activity as determined by M-07e assay, in the culture supernatant of COS7 into which 6GL-TPO, N3/TPO or 09/TPO has been introduced and subsequently expressed. Figure 23 is a graph showing TPO activity, as determined by M-07e assay, in supernatants of COS7 cell cultures in which an insertion derivative or deletion derivative of human TPO has been introduced and expressed. Figure 24 shows an increased platelet count in mice administered TPO via intravenous and subcutaneous injection. Figure 25 shows a dose-dependent increased platelet count in mice after subcutaneous injection of TPO. Figure 26 shows TPO induced increase in platelet count after treatment of mice with 5-FU to induce thrombocytopenia. Figure 27 shows TPO induced increase in platelet count after treatment of mice with nimustine hydrochloride to induce thrombocytopenia. Figure 28 shows TPO induced increase in platelet count in thrombocytopenic mice after bone marrow transplant. Figure 29 shows TPO induced platelet counts after x-ray irradiation of mice to induce thrombocytopenia. Figure 30 shows a dose-dependent increase in platelet count after administration of abbreviated TPO (amino acid 1-163 in SEQ ID NO: 6). Figure 31 shows increase in platelet count after administration of abbreviated TPO (amino acids 1-163 of SEQ ID NO: 6) after administration of nimustine hydrochloride to induce thrombocytopenia. Figure 32 shows that increasing concentrations of Mpl-X added to the human megakaryocyte culture system with increasing block megakaryocyte development. Figure 33 describes TPO activity, which is determined by M-07e analysis of TPO derivatives [Met"2, Lys"<1>, Ala1, Val3, Alg133]TPO(1-163), [Mer2, Lys"<1>, Ala<1>, Val3, Pro<148>]TPO(1-163) and [Mer<2>, Lys"1, Ala1, Val3, Arg115]TPO(1-163) expressed in E. coli. Figure 34 describes TPO activity as determined by M-07e analysis of TPO derivatives [Mer2, Lys"<1>, Ala<1>, Val<3>. Alg<129>]TPO(1-163), [Mer2 Lys"1, Ala<1>, Val3, Arg<143>]T<p>o(i-163), [Mer<2>, Lys"1, Ala1, Val3, leu82]TPO( 1-163), [Met"<2>, Lys<1>, Ala<1>, Val<3>, Leu<146>]TPO(l-163) and [Mer2 Lys"1, Ala1, Val<3>, Arg<59 >]TPO(1-163) expressed in E. coli.

Particularly provided according to the present invention are thrombopoietin (TPO) polypeptides with biological activity that specifically stimulate or increase platelet production and that comprise the amino acid sequence 1 to 332 of SEQ ID NO: 6 or a polypeptide encoded by a polynucleotide that hybridizes under stringent conditions with a polynucleotide encoding the amino acid sequence SEQ ID NO:6. Illustrative polypeptides include those consisting of the amino acid sequences 1 to 163 of SEQ ID NO: 6; 1 to 232 of SEQ ID NO: 6; 1 to 151 of SEQ ID NO: 6; the mature sequence of amino acids of SEQ ID NO: 2,4 and 6; including those having from 1 to 6 NH 2 terminal amino acids removed. Additional illustrative polypeptides of the invention include [Thr<33>, Thr<333>, Ser33* Ile335, Gly336, Tyr33?, Pro338 Tyr339, Asp34<>, Val341, Pro342, Asp343, Tyr344, Ala345, Gly346 Val34? His34», His349, His350, ffis351, His352, His353]TPO, [Asn25, Lys231, Thr333, Ser334, De335, Gly336, Tyr337, Pro338, Tyr339, Asp340, Val34*, Pro342, Asp343, Tyr344, Ala345, Gly346, Val347 , His348, His349, His350, His35<1>, His352 His<353>]TPO, [Asn25]TPO and [Thr33]TPO. Other polypeptides of the invention include derivatives of a TPO polypeptide [AHis<33>]TPO-(1-163), [AArg<ll7>]TPO(1-163), [AGly116]TPO(1-163), [His33, Thr33', Pro34]TPO(l-163), [His33, Ala33', Pro34]TPO(l-163), [His33, Gly33', Pro<34>, Ser<38>]TPO(l-163), [Gly116, Asn11<6>', Arg117]TPO(1-163), [Gly116, Ala116 Arg117]TPO(1-163).

[Gly11<6>, Gly11<6>', Arg117]TPO(1-163), [Ala<1>, Val<3>] TPO (1-163) [Ala<1>, Val<3>, Arg <129>]TPO(1-163), [Ala<1>, Val<3>, Arg133]TPO(1-163), [Ala<1>, Val3, Arg143]TPO(1-163). [Ala<1>, Val<3>, Leu8<2>]TPO(l-163), [Ala<1>, Val<3>, Leu146]TPO( 1-163), [Ala<1>, Val3 , Pro<148>]TPO(l-163), [Ala<1>, Val<3>, Arg59]TPO(l-163), and [Ala<1>, Val3, Arg115]TPO(l-163) , [Arg<129>]TPO (1-163), [Arg<133>]TPO (1-163), [Arg<143>]TPO (1-163), [Leu82]TPO (1-163), [Leu1<46>]TPO (1-163), [Pro<148>]TPO (1-163), [Arg<59>]TPO (1-163) and [Argu5]TPO (1-163).

The TPO polypeptides in the invention also include those which are conventionally bound to a polymer, preferably polyethylene glycol. The TPO polypeptides in the invention may further comprise the amino acids [Met~<2->Lys_<1>], [Met-<1>] or [Gly<1>]. DNA provided in the invention includes those that code for the TPO polypeptide and the derivatives described above and are suitably provided as cDNA, genomic DNA and manufactured DNA.

The invention also provides methods for producing a TPO polypeptide as described above comprising the steps of expressing a polypeptide as encoded by a DNA in the invention in a suitable host, and isolating said TPO polypeptide. Where the TPO polypeptide expressed is a Met"<2->Lys-<1> polypeptide, such methods may further include the step of cleaving Met~<2->Lys_<1> from said isolated TPO polypeptide.

The invention also provides methods for the production of TPO polypeptides such as glutathione-S-transferase (GST) fusion polypeptides. DNA encoding an amino-terminal GST polypeptide, a thrombin recognition peptide and a TPO polypeptide is introduced into a suitable host, the fusion polypeptide is isolated and the GST portion removed by treatment with thrombin. Resulting TPO polypeptides are of the [Gly<1>] structure. It is further provided prokaryotic or eukaryotic host cells transformed or transfected with a DNA sequence according to the invention in a way that enables said host cell to express a polypeptide which has biological activity of specifically stimulating or increasing platelet production.

Pharmaceutical compositions in the invention comprise an effective amount of a TPO polypeptide as mentioned above in combination with a pharmaceutically acceptable carrier and are subject to use in the treatment of platelet disorders, treatment of thrombocytopenia such as that induced by chemotherapy, radiotherapy or bone marrow transplantation. Corresponding treatment methods are also provided in the invention.

The present invention also includes the use of said polypeptides for the production of a drug for the treatment of platelet disorder in a patient. Said platelet disorder may be thrombocytopenia, where said thrombocytopenia may be reduced by chemotherapy, radiotherapy or bone marrow transplantation.

The invention finally provides antibodies which are specifically immunoreactive with TPO polypeptides as described above. Such antibodies are useful in methods for isolating and quantifying TPO polypeptides in the invention.

According to the present invention, a new DNA sequence is provided which codes for a protein that has a TPO activity (hereinafter referred to as "DNA sequence of the present invention"). The DNA sequence in the present invention includes a DNA sequence that codes for the amino acid sequence shown in SEQ ID

. NO: 2,4 or 6 of the sequence listing that follows.

Included in the DNA sequence in the present invention is a DNA sequence that codes for a partially modified (substitution, deletion, insertion or addition) version of the aforementioned amino acid sequence shown in SEQ ID NO: 2, 4 or 6, provided that such modifications do not destroys TPO activity.

In other words, the DNA sequence in the present invention includes DNA sequences that encode protein molecules whose amino acid sequences are mainly amino acid sequences shown in SEQ ID NO: 2, 4 or 6. The term "amino acid sequences are mainly the amino acid sequences shown in SEQ ID NO: 2, 4 or 6" as used here means that said amino acid sequence includes those represented by SEQ ID NO: 2,4 or 6 as well as those represented by SEQ ID NO: 2,4 or 6 that have partial modification therein such as substitution, deletion, insertion, addition or the like, provided that such modifications do not destroy TPO activity.

The DNA sequence in the present invention mainly consists of a DNA sequence that codes for a protein that has a TPO activity.

The words "a DNA sequence encoding the amino acid sequence" include all DNA sequences which may be degenerate into nucleotide sequences.

The DNA sequence in the present invention also includes the following sequences:

(a) DNA sequences represented by SEQ ID NO: 7,194,195 and 196, or complementary strands thereof, (b) DNA sequences that hybridize under stringent conditions to the DNA sequence defined in (a) or fragments thereof, or (c) DNA sequences which hybridises to the DNA sequences as defined in (a) and (b), excluding degeneracy of the genetic code.

Among others, the DNA sequences in the present invention include the following sequences: (a) the DNA sequence which is integrated into the vector pEF18S-A20 (deposition no. FERM BP-4565) carried by an E. coli strain DH5, a vector pHT 1-231 (deposition no. FERM BP-4564) carried by an E. coli strain DH5, a vector pHTFl (deposition no. FERM BP-4617) carried by an E. coli strain DH5, a vector pHGTl (deposition no. FERM BP-4616) carried of an E. coli strain DH5, and which encodes the amino acid sequence of a protein having TPO activity; or (b) DNA sequences which hybridize (under strict conditions) to the DNA sequence defined in (a) or fragments thereof, and which encode the amino acid sequence of a protein having TPO activity.

As used herein, representative "stringent" hybridization conditions are those used in the examples relating to PCR amplification of DNA in the invention by using degenerate and/or unique sequence oligonucleotide primers (probes). See also, for example, Chapters 11 and 14 in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989) and Unit 2.10 in Current Protocols in Molecular Biology, Ausubel et al., Eds., Current Protocols, USA, 1993) .

Also included in the DNA sequence in the present invention is a DNA sequence that codes for a protein that has TPO activity and includes the nucleotide sequence that codes for positions 1-163 in the amino acid sequence represented by SEQ ID NO: 6.

Such DNA sequences can be supplemented with a restriction enzyme cleavage site and/or an additional DNA sequence at the initiation, termination or intermediate site which simplifies construction and rapidly expressed vectors. When a non-mammalian host is used, a preferred codon for gene expression in the host may be incorporated.

An example of the DNA sequence in the present invention is a cDNA molecule that was obtained by preparing mRNA from cells of mammals, including humans, and then screening the cDNA in the usual way from a cDNA library prepared by a known method. Sources of mRNA of this type include columns of a rat hepatocyte-derived cell line McA-RH8994, HTC cells, H4-E-E cell, rat liver, kidney, brain and small intestine human liver and the like.

Another example of the DNA sequence in the present invention is a genomic DNA molecule which is obtained by screening in the usual way from a genomic library prepared by a known method from cells of mammals including humans. Sources of genomic DNA of this type include chromosomal DNA preparations obtained from humans, rats, mice and the like.

A DNA sequence that codes for a TPO derivative can be obtained from the thus obtained cDNA sequence that codes for a protein that has TPO activity, by modifying the cDNA sequence that makes use of known site-directed mutagenesis and thus carrying out partial modification of the corresponding amino acid sequence .

When one has derived the amino acid sequence or the DNA sequence of a protein that has TPO activity in the present invention, a DNA sequence that codes for a partially modified amino acid sequence can be obtained in a simple way by chemical synthesis.

The DNA sequence in the present invention is a useful material for large-scale production of a protein that has TPO activity and makes use of different recombinant DNA techniques.

The DNA sequence of the present invention is useful as a labeled probe for isolating a gene encoding a TPO-related protein, as well as cDNA and genomic DNA encoding TPO or other mammalian species. It is also useful in the gene therapy of human and other mammalian species. In addition, the DNA sequence in the present invention is useful for the development of transgenic mammals that can be used as a eukaryotic host for large-scale production of TPO (Palmiter et al., Science, vol. 222, p. 809-814, 1983).

The present invention also provides a vector in which the aforementioned DNA sequence codes for a protein that has TPO activity is integrated, host cell transformed with said vector and a method for the production of a protein that has TPO activity and which comprises cultivation of said host cells and separation and purification of the expressed protein that has TPO activity.

Examples of host cells useful in this case include cells of prokaryotes such as E. coli and the like, and eukaryotes such as yeast, insects, mammals and the like. Illustrative examples of mammalian cells include COS cells, Chinese hamster ovary (CHO) cells, C-127 cells, baby hamster kidney (BHK) cells, and the like. Illustrative examples of yeast include baker's yeast (Saccharomyces cerevisiae), a methanol assimilating yeast (Pichia pastoris) and the like. Illustrative examples of insect cells include silkworm cultured cells and the like.

With respect to vectors for use in transforming these host cells, pKC30 (Shimatake H. and M. Rosenberg, Nature, 292,128-132,1981), pTrc99A (Amann E. et al., Gene, 69,301-315,1988) or similar is used for the transformation of E. coli cells. For transformation of mammalian cells, pS V2-neo (Southern and Berg, J. Mol. Appl. Genet. 1,327-341,1982), pCAGGS (Niwaet al., Gene, 108, 193-200,1991), pcDL-SR0296 (Takebe et al., Mol. Cell. Biol. 8,466-472,1988) or the like is used. In the case of yeast cells, pG-1 (Schena M. and Yamamoto K.R., Science, 241,965-967,1988) or the like can be used. A transfer vector for recombinant virus construction such as pAc373 (Luckow et al., Bio/Technology, 6,47-55,1988) can be used for transformation of silkworm cells.

If necessary, each of these vectors may contain an origin of replication, selection marker(s), a promoter and the like, as well as an RNA splice site, a polyadenylation signal and the like in the case of vectors for use in eukaryotic cells.

With regard to the origin of replication, a sequence derived from, for example, S V40, adenovirus, bovine papilloma virus or the like can be used in vectors for mammalian cells. A sequence derived from ColEl, R factor, F factor or the like can be used in vectors for E. coli cells. A sequence derived from 2 (im DNA, ARS1 or the like can be used in vectors for yeast cells.

As promoters for gene expression, those derived from, for example, retroviruses, polyoma viruses, adenoviruses, SV40 and the like can be used in vectors for mammalian cells. A promoter derived from bacteriophage', for example, trp, lpp, lac or tac promoter, can be used in vectors for E. coli cells. ADH, PH05, GPD, PGK or MAFO promoter can be used in the vectors for baker's yeast cells, and AOX1 promoter or the like in vectors for methanol assimilating a yeast cell. A promoter derived from a nuclear polyhedrosis virus can be used in vectors for silkworm cells.

Typical examples of selectable markers useful in vectors for mammalian cells

include neomycin (neo) resistance gene, a thymidine kinase (TK) gene, a dihydrofolate reductase gene (DHFR), an E. coli xanthine-guanine phosphoribosyltransferase (Ecogpt) gene and the like. Illustrative examples of selectable markers useful in vectors for E. coli cells include the kanamycin resistance gene, in the case of the ampicillin resistance gene, a terra-

cyclin resistance gene and the like, and those for yeast cells include Leu2, Trpl, Ura3 and the like genes.

Production of protein having TPO activity making use of appropriate combinations of these host-vector systems can be accomplished by transforming appropriate host cells with a recombinant DNA obtained by inserting the gene of the present invention into an appropriate site of the aforementioned vector, culturing the resulting transformant and then separating and purifying the polypeptide of interest from the resulting cells, or culture medium or filtrate. Commonly used ways and methods can be used for this method in combination.

When developing the gene of interest, the original signal sequence may be modified or replaced with a signal sequence derived from another protein to obtain a homogeneous N-terminus of the expressed product. Homogenization of the N-terminus can be carried out by modification (substitution or addition) of the amino acid residues in the N-terminus or its vicinity. In the case of the expression using E. coli as a host cell, for example the lysine residue can be further supplemented in addition to the methionine residue.

The new protein of the present invention that has TPO activity (hereinafter referred to as "the protein of the present invention") comprises proteins that each comprise the amino acid sequence shown in SEQ ID NO: 2, 4 or 6. TPO derivatives whose amino acid sequences are partially modified ( substitution, deletion, insertion or addition) are also included in the present invention, provided that the TPO activity is not destroyed by the modification.

In other words, the proteins of the present invention include protein molecules whose amino acid sequences are mainly the amino acid sequences shown in SEQ ID NO: 2, 4 or 6.

The term "amino acid sequences are essentially amino acid sequences shown in (or represented by) SEQ ID NO: 2,4 or 6" as used means that said amino acid sequences include those shown in SEQ ID NO: 2,4 or 6 as well such as those shown in SEQ ID NO: 2,4 or 6 which have partial modification therein such as substitution, deletion, insertion, addition and the like, provided that such modifications do not destroy the TPO activity.

The protein of the invention includes a protein containing positions 7-151 in the amino acid sequence shown in SEQ ID NO: 6 and which has a TPO activity. Also included in the protein in the present invention is a protein which has TPO activity and which comprises positions 1-163 in the amino acid sequence which is represented by SEQ ID NO: 6.

Examples of other TPO derivatives in the present invention include a derivative whose stability and durability in vivo were improved by the arninoic acid modification (substitution, deletion, insertion or addition), a derivative in which at least one potential glycosylation was changed by deletion or addition, a derivative in which at least one cysteine residue was omitted or substituted with another amino acid residue (alanine or serine residue for example).

The protein in the present invention is preferably characterized in that it is separated and purified from host cells transformed with a recombinant vector containing a cDNA molecule, a genomic DNA molecule or a DNA fragment obtained by chemical synthesis.

When intracellular expression is carried out by using a bacterium such as E. coli as a host, a protein in which an initiation methionine residue is added to the N-terminal side of a protein molecule having TPO activity is obtained, which is also included in the present invention. Depending on the host that is used, the produced protein that has TPO activity may or may not be glycosylated, and all such cases are included in the protein of the present invention.

The protein in the present invention also includes naturally occurring TPO-active proteins purified and isolated from natural sources such as cell culture medium that has TPO activity or human urine, serum or plasma.

A method for purifying TPO from such natural sources is also included in the present invention. Such a purification process can be carried out by using one of or a combination of commonly used protein purification steps such as ion exchange chromatography, lectin affinity chromatography, triazine color affinity chromatography, hydrophobic interaction chromatography, gel filtration chromatography, reverse phase chromatography, heparin affinity chromatography, sulfated gel chromatography, hydroxylapatite chromatography, isoelectric focusing chromatography, metal chelating chromatography, preparative electrophoresis, isoelectric focusing gel electrophoresis and the like. A purification process in which certain techniques are used in combination makes use of physiochemical properties of TPO which can be derived from the examples in the description is also included in the present invention. In addition, an antibody affinity chromatography to which an antibody capable of recognizing TPO can also be used. Furthermore, TPO has been found to be the ligand for Mpl (de Sauvage, et al., Nature 369: 533-538 (1994); Bartley et al., Cell 77: 1117-1124 (1994); Kaushansky et al., Nature 369: 565-568 (1994)), and TPO can be purified using an affinity gel column in which Mpl has been coupled to a resin. More specifically, an example of the column is an Mpl-X column which is prepared by coupling a resin with an extracellular region of Mpl (Mpl-X) prepared by recombinant DNA techniques using CHO cells as a host (Bartley et al. ( 1994), supra).

As described here, the TPO polypeptides in the invention are further characterized by their ability to bind to the Mpl receptor and specifically to its extracellular domain (soluble).

Also included in the present invention is a protein encoded by a DNA part that is complementary to the protein-coding strand of a human cDNA or a genomic DNA sequence of the TPO gene, namely a "complementary inverted protein" as described by Tramontano et al. (Nucl. Acids Res., vol. 12, pp. 5049-5059, 1984).

The present invention also includes a protein from the present invention which is labeled with a detectable marker, for example !25j labeling or biotinylation, and thus makes it possible for a reagent which is useful for the detection and quantification of TPO or TPO receptor-expressing cells in fixed samples such as tissue and liquid samples such as blood, urine and the like.

Biotinylated protein of the present invention is useful in the case of binding to immobilized streptavidin to remove megacryoblasts from bone marrow at the time of

autogenous bone marrow transplantation. It is also useful in the case of binding to immobilized streptavidin to concentrate autogenous or allogeneic megakaryocytic columns at the time of autogenous or allogeneic bone marrow transplantation. A conjugate of TPO with a toxin such as ricin, diphtheria toxin or the like or with a radioactive isotope is useful in antitumor therapy and in conditioning for bone marrow transplantation.

The present invention also provides a nucleic acid material useful when labeled with a detectable marker including a radioactive marker or a non-radioactive marker such as biotin or when used in a hybridization procedure for detecting the position of the human TPO gene and/or its related gene family on a chromosomal map. Such material is also useful in the confirmation of human TPO gene disorders at the DNA level and can be used as a genetic marker for the confirmation of adjacent genes and their disorders.

The present invention also includes a pharmaceutical composition containing a therapeutically effective amount of protein of the present invention together with a useful and effective diluent, antiseptic, solubilizer, emulsifier, adjuvant and/or carrier. The term "therapeutically effective amount" as used herein means an amount that provides a therapeutic effect for the specified condition and route of administration to be used. Such a composition is used in the form of a liquid, freeze-dried or dried preparation and comprises a diluent selected from different buffers (Tris-HCl, acetate and phosphate for example) with different pH values and ionic strengths, a surface adsorbing barrier additive such as albumin or gelatin, a surfactant such as Tween 20, Tween 80, Pluronic F68 and bile acid salt, a solvent such as glycerol or polyethylene glycol, an antioxidant such as ascorbic acid or sodium metabisulfite, an antiseptic such as thimerosal, benzyl alcohol or paraben and a carrier or tonicity agent such as lactose or mannitol. A composition was also used which comprises covalent binding of the protein from the present invention to a polymer such as polyethylene glycol, gelation of the protein with the metal ion, incorporation of the protein in a granular preparation, or on the surface, consisting of a polymer compound such as polylactic acid, polyglycolic acid, or hydrogel or incorporations of protein into liposomes, microemulsion, micelles, single or multi-layered vesicles, red cells or spheroplasts. Such a composition will exert influence on physical conditions, solubility, stability, in vivo release rate and in vivo clearance of TPO. Choice of composition can be decided depending on the physical and chemical properties and the TPO-active protein used. The present invention also includes a granular composition in which granules are coated with a polymer such as poloxamer or poloxamine and TPO which binds to antibodies for tissue-specific receptors, ligands or antigens or to tissue-specific receptor ligands. Other examples of the composition in the present invention are those that are in the form of granules, and have a protective coating and contain a protease inhibitor or a penetration enhancer, for use in different administration routes such as parenteral, pulmonary, transnasal and oral administration.

The pharmaceutical composition comprising the protein of the present invention can be administered several times per day, usually in an amount of 0.05 mg to 1 mg/kg (as TPO protein) of body weight depending on conditions and the patient's gender, route of administration and the like.

The pharmaceutical composition containing the protein of the present invention can be administered at a dose of 25,000-500,000,000 of active ingredient (on the basis of relative activity by M-07e analysis which is described later) per kg of body weight one to several times per day and 1 to 7 days per week depending on the symptom, gender and route of administration.

The present invention has confirmed that with regard to the C-terminal site of human TPO, the activity is maintained even when the amino acid residues up to the 152nd amino acid sequence shown in SEQ ID NO: 6 are omitted and that with regard to the N-terminal site, the activity is maintained even when the amino acid residues up to the 6th amino acid sequence are omitted.

The protein which has TPO activity containing the 7th to 151st amino acid sequence of SEQ ID NO: 6 and which is modified (substitution, deletion, insertion or addition) or other parts can preferably be used as an effective ingredient in the present invention. More preferred TPO derivative is that which has the 1st to 163rd amino acid sequence of

SEQ ID NO: 6.

A composition containing at least one of additional hematopoietic factors such as EPO, G-CSF, GM-CSF, M-CSF, IL-1, EL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, EL-10, IL-11, IL-12, IL-13, LIF and SCF, in addition to the proteins of the following invention, are

. also included in the present invention.

The protein of the following invention is useful for the treatment of various thrombocytopenic diseases, either alone or in combination with other additional hematopoietic factors. Examples of other hematopoietic factors include EPO, G-CSF, GM-CSF, M-CSF, IL-1, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL- 8, IL-9, IL-10, IL-11, IL-12, EL-13, LIF and SCF.

There are many diseases that are characterized by platelet disorders such as thrombocytopenia due to damage in the production of platelets or reduction in the lifespan of platelets (caused by the destruction or consumption of platelets), which can be treated with the protein of the present invention. For example, it can be used to improve platelet recovery in a thrombocytopenic patient due to congenital Fanconi anemia or aplastic anemia caused by chemotherapy or radiation, myelodysplastic syndrome, acute myelocytic leukemia, aplastic crisis after bone marrow transplantation. It is also useful in the treatment of thrombocytopenia due to reduced production of TPO. Thrombocytopenia due to a reduction in platelet or megakaryocyte lifespan includes idiopathic thrombocytopenia purpura, hypoplastic anemia, acquired immunodeficiency syndrome (AIDS), disseminated intravascular coagulation syndrome, and thrombotic thrombocytopenia. In addition, it is also useful in this case of autotransfusion of platelets, where TPO is administered to a patient prior to surgical operation to increase the number of the patient's own platelets and the thus increased platelets are used as transfusion platelets at the time of surgery. Another application of the protein in the present invention is the treatment of diseases which are the result of temporal omission or damage of platelets caused by other chemical or pharmaceutical drugs or healing treatment. TPO can be used with the caveat of improving the release of new "intact" platelets in such patients.

The present invention further relates to an isolated antibody specific for TPO. The protein in the present invention can be used as antigens, and corresponding antibodies include both monoclonal and polyclonal antibodies and chimeric antibodies, namely "recombinant" antibodies, produced by conventional methods. To produce such an anti-TPO antibody, human TPO itself can be used as the antigen, or alternatively a partial peptide of human TPO can be used as antigen. Where an antibody to the antigen of such a peptide is used, the epitope can be defined by specifying the antigenic region. On the other hand, where an antibody to the antigen protein (TPO) itself is utilized, the antigenic region can be clarified by analyzing the epitope of the antibody. In such cases, these antibodies, each of which has thus prepared an epitope, can be utilized to fractionate, detect, quantify and purify different TPOs that differ in their properties such as type of added sugar chain, length of peptide chains, etc.

A TPO peptide containing the thus selected amino acid sequence is synthesized and bound to a suitable carrier protein, for example albumin, KLH (keyhole limpe-themocyanin) or the like, by covalent binding to produce an immunoreactive antigen. Alternatively, a multiple antigen peptide (MAP) type peptide is prepared and is an antigen, according to the method of Tam (Proe. Nati. Acad. Sei. (USA) 85:5409-5413, 1988). With the antigen produced in this way together with an adjuvant or the like, mammals, fowls etc. which are generally utilized to make them produce antibodies, such as rabbits, mice, rats, goats, sheep, chickens, hamsters, horses, guinea pigs, are immunized. etc. From the thus-immunized animals, their anti-serum and cells were recovered, from which a polyclonal antibody is obtained. Where a peptide is used as an antigen, the epitope can be defined by specifying the antigenic region. In the event, regions of different TPO molecules having different molecular weights are screened and identified by immunological engineering. As an example, a peptide-skipped region can be screened and identified. In addition, an antibody gene or part of the gene is cloned from cells expressing the intended antibody to obtain an antibody molecule that has been expressed by genetic engineering.

In the context of polyclonal antibody that generally contains different antibodies that are specific for different antigenic determinants (epitope), a monoclonal antibody is specific for a single antigenic determinant on antigens. A TPO-specific antibody is useful for improving selectivity and specificity in an antigen-antibody reaction-assisted diagnosis and an analytical analysis method and when carrying out the separation and purification of TPO. In addition, such antibodies can be used to neutralize or remove TPO from serum. Monoclonal antibodies are also useful for the detection and quantification of TPO, for example in serum from whole blood.

The anti-human TPO antibody in the present invention can be used as a ligand in affinity chromatography for the purification and isolation of human TPO. To fix the antibody to make it useful in such affinity chromatography, any conventional method of fixing various enzymes can be used. For example, a method can be used by using a carrier of CNMr-activated sepharose 4B (manufactured by Pharmacia Fine Chemicals Co.) or the like. To truly purify human TPO using fixed anti-human TPO antibody, fixed anti-human TPO antibodies are loaded into a column and a human TPO-containing liquid is passed through the column. In this operation, a large amount of human TPO is adsorbed onto the support in the column. As a solvent for elution, for example glycine-HC1 buffer (pH 2.5), sodium chloride solution, propionic acid, dioxane, ethylene glycol, chaotropic salt, guanidine hydrochloride, urea etc. can be used. When elution with such a solvent, human TPO is eluted with a high purity.

The antibody in the present invention is used for the determination of human TPO by immunochemical quantitative analysis, particularly by enzyme immunoassay which is carried out in a solid phase sandwich process.

An advantage of monoclonal antibodies is that they can be produced by hybridoma cells in a medium that does not contain any other immunoglobulin molecules. Monoclonal antibodies can be prepared from culture supernatants of hybridoma cells or from mouse ascites induced by intraperitoneal injection of hybridoma cells. The hybridoma technique originally described by Kohler and Milstein (Eur. J. Immunol. 6:511-519,1976) can be used to generate hybrid cell lines that contain high levels of monoclonal antibodies to a number of specific antigens.

In order to isolate the antibody in question from the antibody-containing material thus obtained, such as anti-serum, etc. when purifying this, one or more steps (affinity chromatography such as protein A affinity chromatography, protein G affinity chromatography, avidgel chromatography, anti- immunoglobulin fixed gel chromatography, etc; as well as cation exchange chromatography, anion exchange chromatography, lectin affinity chromatography, dye adsorption chromatography, hydrophobic interaction chromatography, gel permeation chromatography, reverse phase chromatography, hydroxylapatite chromatography, fluoroapatite chromatography, metal chelating chromatography, isoelectric point chromatography, partition electrophoresis, isoelectric point electrophoresis, etc.) which can generally be used for protein purification. Apart from these, a method for antigen affinity purification can also be used, where a gel carrier or membrane, to which is chemically bound a human TPO protein itself or a peptide containing the region of the antigen or part of the region, or which is a molecule which has the ability to recognize the antibody in question, is produced, an antibody-containing material is added to this so that the antibody in question is made and adsorbed on the carrier or membrane, and the thus adsorbed antibody is then eluted and recovered under suitable conditions.

The present invention also allows the use of endogenous DNA sequences that code for a TPO polypeptide for the production of large quantities of polypeptide products. For example, in vitro and ex vivo homologous recombination procedures can be used to transfect host cells to provide expression or enhanced expression of polypeptides. Preferred host cells include human cells (for example, liver, bone marrow and the like) where a promoter or enhancer sequence is inserted using flanking sequences that are homologous to a target region in a cellular genome but results in TPO polypeptide expression being carried out or enhanced. See, for example, US-PS 5,272,071, published PCT WO 90/14092, WO 91/06666 and WO 91/09955.

The present invention also provides methods for producing a TPO polypeptide as described above comprising the step of expressing a polypeptide encoded by a DNA from the invention in a suitable host, and isolating said TPO polypeptide. Where the TPO polypeptide as expressed is a Met'<2->Lys"<1> polypeptide, such methods can further include the step of cleaving Mef^-Lys"1 from said isolated TPO polypeptide.

Methods are also provided for producing TPO polypeptides having [Gly<1>] structure and comprising the step of introducing into a suitable host cell a DNA encoding a glutathione-S-transferase (GST) polypeptide 5' to the TPO polypeptide coding sequences, GST and TPO polypeptide coding sequences separated by DNA encoding a thrombin recognition polypeptide, isolation of GST-TPO expression product and treatment of the expressed polypeptide with thrombin to remove GST amino acids. Resulting TPO polypeptides have a [Gly-<1>] structure.

The following describes the present invention in detail.

Purification of rat TPO. analysis of partial amino acid sequences of purified rat TPO and analysis of biological characteristics of purified rat TPO

The present invention has made the first attempt to purify a protein (rat TPO) which has an activity to improve the proliferation and differentiation of rat CFU-MK. In this purification study, many trials and errors were made on the purification procedure, such as the selection of different natural sources, the choice of gels for chromatography and separation methods. As a result, the present invention has succeeded in purifying a protein having a TPO activity from the blood plasma of thrombocytopenic rats induced for X-ray or gamma irradiation, and makes use of TPO activity as a marker based on a rat CFU-MK analysis which will be described in the following in (reference example), and determination of partial amino acid sequences of purified protein (examples 1 and 2).

Biological characteristics of plasma-derived rat TPO were also investigated (Example 3).

An outline of the procedures from purification of rat TPO to determination of partial amino acid sequences of purified protein is shown below. (i) Blood plasma sample was prepared from approximately 1,100 thrombocytopenic rats induced by X-ray or gamma irradiation and subjected to a Sephadex G-25 chromatography, an anion exchange chromatography (Q-Sepharose FF) and a lectin chromatography (WGA-agarose) in that order, and thus obtain a WGA-agarose-adsorbed TPO-active fraction. (ii) Subsequently, the WGA-agarose-adsorbed TPO-active fraction thus obtained was subjected to a triazine dye affinity chromatography (TSK AF-BLUE 650MH), a hydrophobic interaction chromatography (Phenyl Sepharose 6 FF/LS) and a gel filtration chromatography (Sephacryl S-200 HR ) in that order. Since the TPO activity was divided into four peaks (F1 as the highest molecular weight fraction, followed by F2, F3 and F4) by Sephacryl S-200 HR gel filtration, each of the TPO active fractions F2 and F3 was concentrated to obtain a high molecular weight TPO samples F2 and a low molecular weight TPO sample F3 which was used separately in subsequent purification steps. (iii) The low molecular weight TPO sample F3 was subjected to a preparative reverse phase chromatography (YMC-Pack PROTEIN-RP), a reverse phase chromatography (YMS-Pack CN-AP) and another reverse phase chromatography (Capcell Pack Cl) in that order. When the TPO-active fraction thus obtained was applied to sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (SDS-PAGE) and TPO-active substance was extracted from the gel, the presence of TPO activity was confirmed in a band corresponding to molecular weights of about 17,000 to 19,000 under no - mitigating conditions. (iv) As a consequence, all TPO-active fractions were applied to SDS-PAGE under non-reducing conditions and transferred to a PVDF membrane. By carrying out systematic limited enzymatic hydrolysis of the protein on the PVDF membrane into peptide fragments, partial amino acid sequences of rat TPO protein were determined. Based on amino acid sequence information of two peptide fragments, cloning of the rat TPO gene was carried out. (v) Separately from this, high molecular weight TPO sample F2 obtained by Sephacryl S-200 HR was subjected to purification in the same way as in the case of low molecular weight TPO sample F3. When a TPO active fraction obtained in the last step of reverse phase chromatography (Capcell Pack Cl) was applied to SDS-PAGE and TPO-active substance was extracted from the gel, the presence of TPO activity was confirmed in a band corresponding to a molecular weight of approx. 17000 to 22000 under non-reducing conditions.

Specialization of rat TPO-producing cells,

production of mRNA and construction of rat cDNA library Since the thus obtained rat TPO has been derived from blood plasma, it was necessary to screen organs or cells as a source of mRNA for use in cDNA cloning. As a consequence, TPO activity in different organs and cell culture supernatants was screened based on biochemical and biological properties of rat plasma-derived TPO. As a result, TPO activities almost similar to rat plasma-derived TPO were found in the culture supernatant of rat hepatocyte-derived cell lines McA-RH8994, HTC and H4-II-E, as well as in the culture supernatant of rat primary hematocytes (Example 4).

On the other hand, expression vector pEF18S was constructed from 2 expression vectors pME18S and pEFBOS (Example 5). Construction of this vector was made possible by simple cloning of the cDNA using a high-throughput expression vector that has many cloning sites that can be used for integration of inserts.

In addition to the expression vector above, pUC and pBR based plasmid vectors and phage vectors are mainly used for the construction of cDNA libraries. McA-RH8994 cells were cultured, homogenized by addition of guanidine thiocyanato solution and they were then subjected to CsCl tetmets gradient centrifugation to obtain total RNA (Example 6).

Production of RNA can also be done by a hot phenol method, an acidguanidinium phenol chloroform method and the like.

After purification of poly (A)<+> RNA from total RNA using oligo dT-immobilized latex particles, first strand cDNA was synthesized by the action of reverse transcriptase using oligo dT as a primer to which a restriction enzyme Noti recognition sequence had been added added, followed by synthesis of second-strand cDNA using RNase H and E. coli DNA polymerase I. An EcoRI adapter was added to the double-stranded cDNA thus obtained, and the resulting cDNA was ligated with a mammalian cell expression vector pEF18S constructed in Example 5, which has been cleaved with Noti and EcoRI and the thus ligated cDNA was transferred into competent cells of an E. coli strain DH5 to construct a cDNA library (Example 7).

(C) Preparation (cloning) of rat TPO cDNA fragment by PCR

A DNA sequence was deduced from the partial amino acid sequence of rat TPO which

had been purified from rat blood plasma, to synthesize degenerate primers to be used in polymerase chain reaction (PCR). Primers to be used can also be obtained based on an amino acid sequence different from the position of the primers used here. Highly degenerate primers can also be used without using inosine. In addition, primers with reduced degeneracy can be designed to make use of a codon that has a high utilization in the rat (Wada et al., Nucleic Acids Res., vol. 18, pp. 2367-2411, 1990).

When plasmid DNA was extracted from the entire part of the cDNA library prepared above and PCR was carried out using the extracted DNA as a template, a band of about 330 bp was detected and which was then determined by nucleotide sequence analysis as a DNA fragment (A1 fragment) that codes for a portion of rat TPO (Example 8).

(D) Screening of rat TPO cDNA by PCR, sequence of rat

TPO cDNA and confirmation of TPO activity

The cDNA library prepared above was divided into poles, each containing about 10,000 clones, and plasmid DNA was extracted from each 100 poles. When PCR was performed using each pole of plasmid DNA as template and primers newly synthesized based on the nucleotide sequence of the A1 fragment, bands that were considered to be specific were detected in three poles. One of these three poles was divided into sub-poles, each containing about 900 clones, and

Plasmid DNA was purified from 100 subpools by performing PCR in the same manner. As a result, a specific band was detected in three sub-poles. One of these pools was divided into sub-pools each containing 40 clones, and finally each clone was screened by PCR in the same way. As a result, a clone pEF18S-A26 which appeared to encode rat TPO cDNA was isolated (Examples 9 and 10).

When the nucleotide sequence of this clone was analyzed that encoded the partial amino acid sequence of protein purified from rat blood plasma, thus confirming a strong possibility that this clone contains rat TPO cDNA (Example 10).

When plasmid DNA was purified from clone pEF18S-A26 obtained above and transfected into COS 1 cells, TPO activity was found in the culture supernatant of the transfected cells. As a result, clone pEF18S-A26 was confirmed to contain a cDNA encoding rat TPO (Example 11).

(E) Detection of TPO mRNA in various rat tissues

TPO mRNA expression in rat tissue was analyzed by PCR, and specific expression was

detected in brain, liver, small intestine and kidney (Example 12).

(f) Construction of human cDNA library

Based on the results of Examples 4 and 12, liver was selected as the starting material for cloning human TPO cDNA. Consequently, a cDNA library was constructed using a commercial mRNA derived from normal human liver. Using pEF-18S as a vector in the case of the rat library, cDNA was synthesized by the same method as the case of the rat to obtain a cloned cDNA library using the restriction enzyme Noti and EcoRI. The library produced by introducing vector-ligated cDNA into E. coli DH5 contained about 1,200,000 clones (Example 13).

(G) Preparation (cloning) of human TPO cDNA fragment by PCR

Several primers for PCR were synthesized based on the nucleotide sequence of clone pEF18S-A20 encoding rat TPO cDNA. When cDNA was synthesized using a commercial mRNA derived from normal human liver and PCR was carried out using these primers and cDNA as a template, a band around 620 bp was observed. When the nucleotide sequence was analyzed, it appeared that this clone contains a DNA fragment that has a homology of about 86% with rat TPO cDNA, and thus confirms a strong possibility that this is part of a gene that codes for human TPO (Example 14 ).

(H) Screening of human TPO cDNA by PCR, sequence of

human TPO cDNA and confirmation of TPO activity

The human cDNA library prepared above was amplified, divided into two poles each containing about 100,000 clones, and plasmid DNA was extracted from each of 90 poles. When PCR was carried out using each pole of plasmid DNA molecules as a template and primers newly synthesized based on the nucleotide sequence of the human TPO fragment obtained in example 14, possible bands were detected in three poles. One of these pools was divided into sub-pools each containing 5000 clones and plasmid DNA was purified from each of 90 sub-pools. When PCR was carried out using each pole plasmid DNA as a template in the same way, possible bands were detected in 5 sub-poles. When one of these pools was divided into sub-pools each containing 250 clones and plasmid DNA was purified from each of 90 sub-pools and subjected to PCR, possible bands were detected in three sub-pools. When one of these pools was divided into sub-pools each containing 30 clones and plasmid DNA was purified from each of 90 sub-pools and subjected to PCR, possible bands were detected in three sub-pools. Then 90 colonies were isolated from one of these sub-poles and plasmid DNA purified from each of these colonies was subjected to PCR to finally obtain a clone named HL34 (Example 15).

When the nucleotide sequence of the plasmid DNA in this clone was analyzed, it appeared that this clone contains a cDNA which has approximately 84% homology with the rat TPO cDNA nucleotide sequence (Example 16).

When the cloned plasmid DNA was purified and transfected into COS 1 cells, the TPO activity was found in the culture supernatant of the transfected cells. As a result, it was confirmed that this plasmid clone contains cDNA encoding human TPO (Example 17).

However, this cDNA appears to be an artificial product of cloning, because no termination codon was found in this clone and a poly A tail-like sequence was found at the 3' end. Consequently, an expression vector encoding the amino acid sequence excluding a portion corresponding to the poly A tail-like sequence was constructed. When the vector thus constructed was expressed in COS 1 cells, TPO activity was found in the resulting culture supernatant (Example 18).

DNA fragment in the 3'-end region of human TPO was obtained by PCR to analyze the structure of full-length cDNA.

Determination of the nucleotide sequence of this fragment showed that it partially overlaps with the cDNA carried by clone HL34 obtained in Example 15. It was also expected that a full-length human TPO cDNA may contain an open reading frame thereof and code for a protein consisting of 353 amino acids. It was thus suggested that human TPO consisted of 350 amino acids including a signal sequence of 21 amino acids (Example 19).

In addition to the above cloning method, cloning of human TPO cDNA can be achieved by colony hybridization or plaque hybridization by using a rat TPO cDNA fragment as a probe, by using a library constructed by using a pUC or pBR based plasmid vector, a 'fag based vector or similar. When designing a degenerating probe, inosine can be used to reduce the degree of degeneration. When an analysis method that can detect TPO activity in a specific way or with a high sensitivity is available, it is possible to carry out expression cloning by using an expression library as was used in the present invention.

However, since the content of human TPO-encoding RNA seems to be extremely low in normal human liver, where this is described later in example 15 (content calculated from the results of this example was in a ratio of 1 to 3 per million), hybridization screening there a synthesized oligonucleotide or a rat or human TPO cDNA fragment used as a probe will be extremely difficult to implement, because the number of clones or plaques to be processed becomes bulky and the sensitivity and specificity of the hybridization method is detrimental to those of the PCR method. In connection with the present invention, colony hybridization has actually been carried out on 2 million clones in the normal human liver cDNA library prepared in example 13 by using a rat TPO cDNA fragment as a probe, but a human TPO cDNA clone was unable to be obtained.

(I) Reconstruction of human normal liver-derived cDNA

Since clone HL34 obtained in Example 15 appeared to contain an incomplete cDNA, a cDNA library was reconstructed using a commercial normal human liver-derived poly (A)<+> RNA preparation to obtain a complete human TPO cDNA. The library (hTPO-F1) prepared by introducing vector-ligated cDNA into E. coli DH5 contained 1.0 x 10<6> transformants (Example 20).

(J) Screening of TPO cDNA clone, sequence determination and expression of human TPO cDNA. and confirmation of TPO activity

Primers for PCR were synthesized based on the nucleotide sequence (SEQ DD NO: 3) of a partial cDNA obtained in Example 14 and the nucleotide sequence (SEQ ID NO: 196) from a complete cDNA from human TPO as indicated in Example 19.

Human liver cDNA library (hTPO-Fl) constructed in Example 20 was divided into three poles (pole #1-3). PCR was carried out using plasmid DNA prepared from each pole as a template and synthesized primers. As a result, a DNA fragment having an expected size was amplified when plasmid DNA prepared from pol #3 was used. The #3 pool was then divided into sub-pools each containing 15,000 transformants and screening using PCR was carried out as above. As a result, amplification of DNA having the expected size was found in 6 out of 900 poles. When one of these positive poles divided into subpoles, each containing 1000 clones and plasmid DNA was extracted and PCR was carried out in the same way, DNA amplification was not observed. This is considered to be caused by a low recovery of plasmid DNA which is due to poorer growth of the clone of interest compared to other clones. Originally #3 pole was therefore spread on 100 LB plates at such an inoculation size that 4100 colonies grew on each LB plate, and in a replicate plate was prepared at each of the thus inoculated plates. When PCR using these DNA extracts from colonies that grew on the plate was carried out in the same manner as described above, amplification of a band was observed as expected in one of 100 subpoles.

Two replicate filters were prepared from the plate from this subpole, and colony hybridization was performed using the labeled probe (EcoRI/BamHI fragment of plasmid pEF18S-HL34). As a result, a single signal considered to be positive was observed. Colonies were collected from the original plate and plated again on an LB plate. DNA samples were prepared from a total of 50 colonies grown on the plate and subjected to PCR to finally obtain a clone designated pHTF1 (Example 21). When screening the cDNA clone, methods such as hybridization, PCR and the expression clone were mainly used and the combination of these methods increased the efficiency and sensitivity of the screening, or reduced the work. When the nucleotide sequence of the thus obtained clone pHTF1 was determined, it was found that the clone pHTF1 had an open reading frame and the amino acid sequence of the protein is considered to be encoded by an open reading frame which completely coincides with the deduced amino acid sequence (SEQ DD NO: 6) of human TPO . The nucleotide sequence differed from the deduced one (SEQ ID NO: 196) in 3 positions, but these differences caused no amino acid yield. It is confirmed that human TPO protein comprises 353 amino acid sites which have a 21 amino acid signal sequence. When plasmid DNA is prepared from the thus obtained clone pHTF1 and transfected with COSI cells, TPO activity was found in the culture supernatant (Example 23).

(K) Screening of human TPO chromosomal DNA by plaque reduction,

sequencing and expression of human TPO

chromosomal DNA and confirmation of TPO activity

A genomic library provided by Prof. T. Yamamoto at the Gene Research Center, Tohoku University was inoculated onto 19 NZYM plates in such a seeding size that one plate contained 30,000 phage particles, and two replicate filters were prepared at each of the 18 plates thus prepared. A human TPO cDNA fragment (base position number 178 to 1025 in SEQ ID NO: 7) was amplified by PCR and purified. By using the purified fragment labeled with <32>p as a probe, plaque hybridization was carried out. As a result, 13 positive signals were obtained. Plaques were collected from the original plates and re-inoculated onto NZYM plates at such a seeding size of 1000 plaques were formed on one plate. Two replicate filters were prepared from each of the resulting plates by performing plaque hybridization under the same conditions as described above. The result was that positive signals were detected on all filters of the 13 groups. A single plaque was recovered from each of the resulting plates to prepare phage DNA. Phage DNA samples prepared in this way from 13 clones were examined for the presence of the cDNA coding region by PCR. 5 of the 13 clones appeared to contain the entire amino acid coding region expected from cDNA. Consequently, one of these clones (clone 6HGT1) was selected and analyzed by Southern blotting (using the aforementioned probe). Since a single band of 10 kb was observed in the case of Hindin cleavage, DNA from clone OHGT1 was digested with Hindm and subjected to agarose gel electrophoresis. The 10 kb band was excised from the gel, purified and subcloned into a cloning vector pUC13 to finally obtain a clone named pHGT1 (Example 24).

When the nucleotide sequence of the clone pHGT1 thus obtained was determined, the DNA carried by the clone was found to contain an entire coding region of the protein as expected in Example 19, and the nucleotide sequence of the coding region completely coincides with that of the expected nucleotide sequence (SEQ TD NO : 196).

In addition, a region corresponding to amino acid-coding exon 4 contained intron and the nucleotide sequence was different from that in complete cDNA carried by clone pHTF1 obtained in Example 21 (SEQ ID NO: 7).

When the nucleotide sequence of the four remaining clones among 5 chromosomal DNA

clones independently selected by screening were determined, two out of four clones were found to be identical to clone pHGT1 and the other two clones were the same as clone pHGT1 except that they had a different nucleotide in the 3' non-coding region as that observed with clone pHTF1 (Example 25).

An EcoRI fragment in the obtained plasmid clone pHGT1 was ligated with an expression vector PEF18S to obtain human TPO expression plasmid pEFHGTE. When plasmid DNA (pEFHGTE) was prepared and transfected into COSI cells, TPO activity was found in the culture supernatant (Example 26).

(L) Preparation of human TPO deletion derivatives, expression in COSI cells

and confirmation of TPO activity

The results from examples 18 and 22 showed that human TPO protein could exhibit biological activity even after removal of its carboxy-terminal end. Thus, to further analyze biologically active proteins, deletion derivative experiments were carried out. A series of expression plasmids were prepared by PCR using DNA of plasmid clone pHT 1-231 obtained in Example 18 as a template and synthesized oligonucleotides as primers.

Expression plasmids containing DNA encoding human TPO deletion derivatives lacking the carboxyl-terminal region of the TPO protein, i.e. deletion derivatives encoding positions 1-211, 1-191, 1-171 and 1-163 amino acid sequence, were obtained. When plasmid DNA was prepared from each of the clones and transfected into COSI cells, TPO activity was detected in each of the culture supernatants (Example 27).

When derivatives that have a series of deletions of C-terminal amino acid sites up to position 151 and other derivatives that have respective omissions of N-terminal side amino acid residues up to position 6,7 and 12 were designed and activity after expression in COSI cells was measured for to analyze the TPO activity-bearing region further in detail, and the activity became undetectable when N-terminal side amino acid residues are omitted up to position 7 or C-terminal side amino acid residues are omitted up to position 151.

(Examples 28 and 29).

Derivatives of a protein having TPO biological activity can be obtained by modification (deletion, substitution, insertion or addition) of the cDNA encoding the protein. Methods such as PCR, site-directed mutagenesis and chemical synthesis can be used for modification.

(M) Expression of human TPO cDNA in CHO cells and purification of TPO

An expression vector pHTP1 was constructed and this can express cDNA encoding a deduced amino acid sequence of human TPO as shown in SEQ ID NO: 6, in animal cells. (Example 30).

The TPO cDNA region in pHTP1 was used to construct a vector pDEF202-hTPO-P1 for its expression in CHO cells. (Example 31).

As a result of transfection of the vector in CHO cells and subsequent selection of transfected cells, transfected cells in which the expression vector carrying human TPO cDNA has been integrated into the chromosome of CHO were obtained. (Example 32).

In example 32, large-scale cultivation was carried out on a human TPO-producing CHO cell line (CHO 28-30 cell, resistant to 25 nM MTX) which is produced by transfection of human TPO expression plasmid pDEF202-hTPO-P1 in CHO cells. (Example 55).

From the culture supernatant, 100 L, human TPO was purified. (Example 56).

Alternatively, human TPO was purified from the culture supernatant of Example 55 in a different manner. (Example 57).

(N) Expression of human TPO cDNA in X63. 6. 5. 3. cells and confirmation of

Activity

Using the TPOm cDNA region encoded in plasmid pBLTEN prepared in Example 30, an expression vector BMCGSneo-hTPO-P1 was constructed for use in X63.6.5.3. cells. (Example 33).

When X63.6.5.3. cells were transfected with this vector, a transformant cell was obtained in which the human cDNA encoding expression vector was integrated into the chromosome. When this cell was cultured, the TPO activity was detected in the culture supernatant. (Example 34.

(O) Storm<g>de expression of human TPO in COS 1 cells, and purification,

molecular weight measurement and biological characteristics thereof The expression vector pHTP1 prepared in example 30 was transfected into COS 1 cells to obtain a large amount (in total about 40 litres) of culture supernatant containing expression products. (Example 35).

Purification of TPO was carried out from about 7 liters of COS 1 cell serum-free culture supernatant prepared in example 35 containing the expression vector pHTP1-derived TPO. It was possible to obtain high activity TPO through the step of a hydrophobic interaction chromatography, a cation exchange column chromatography, a WGA column chromatography and a reverse phase column chromatography. (Example 36).

TPO partially purified from the COS 1 culture supernatant was subjected to molecular weight measurement and biological characterization analysis. (Examples 37 and 38).

(P) Expression of human TPO in E. coli

A vector pGEX-2T/hT(l-174) was constructed for use in expression of a fusion protein (referred to as "GST-TPO(l-174)") of glutathione-S-transferase and human TPO (amino acid residues in the positions 1 to 174) in E. coli. In this case, part of the human TPO cDNA nucleotide sequence (about half of the area on the 5' side) was replaced with the E. coli preference codon. (Example 39).

GST-TPO(1-174) is expressed in E. coli, and the resulting cells were disintegrated and then GST-TPO(1-174) contained in the precipitated fraction was made soluble.

As a result of combination of different steps including examination of TPO refolding conditions, examination of purification conditions (glutathione affinity column, cation exchange column and the like) and cleavage of GST protein region with thrombin, it was possible to carry out partial purification of a protein containing a TPO amino acid sequence that formed . It was confirmed that this protein shows TPO activity in the rat CFU/MK assay system. (Examples 40 and 41).

A vector pCFM536/h6T(l-163) was also constructed for use in expression in E. coli of a mutant type of human TPO protein (referred to here as "h6T(l-163)")" where in the 1-position is Ser residue and 3-position Ala residue in human TPO (amino acid residues in positions 1 to 163) respectively replaced with an Ala residue and a Val residue, and at the same time a Lys residue and a Met residue were respectively added to -1 and -2 positions. The whole part of the human TPO cDNA nucleotide sequence (corresponding to 1 to 163 amino acid residues) carried by this vector was replaced by the E. coli preference codon.

(Example 42).

h6T(1-163) was expressed in E. coli, the resulting cells were lysed and then the solution of h6T(1-163) contained in the precipitated fraction and the conditions for refolding the proteins were investigated, thus succeeding in partial purification of a protein containing a TPO amino acid sequence as designed. It was confirmed that the protein shows TPO activity in the rat CFU-MK assay system. (Examples 43 and 44).

In addition, a vector pCFM536/hMKT(l-163) was constructed for use in expression in E. coli of a mutant type human TPO protein (which was referred to as "hMKT(l-163)") in which a Lys residue and a Met residue respectively, -1 and -2 positions were added. hMKT(1-163) was expressed in E. coli in the same way as described in example 43, and the expression protein was subjected to SDS-PAGE, transferred onto a PVDF membrane and then subjected to N-terminal amino acid sequence analysis, thus confirming that this the protein contains the designed amino acid sequence. (Example 52).

In addition, the vector pCFM536/hMKT(l-332) was used for expression in E. coli of a mutation-type human TPO protein in which a Lys residue was added in position -1 and a Met residue in position -2 and human TPO (amino acid positions 1 to 332) (referred to as "hMKT (1-332)") constructed. hMKT (1-332) was expressed in E. coli in the same manner as in Example 42, and its expression was confirmed by Western blotting using anti-human TPO peptide antibody prepared in Example 45 as indicated below.

(Q) Preparation of anti-TPO peptide antibody and anti-TPO peptide

antibody column

Rabbit polyclonal anti-TPO peptide antibodies were prepared by synthesized peptides corresponding to the three partial regions of the rat TPO amino acid sequence determined in Example 10. It was confirmed that these antibodies can recognize rat and human TPO molecules. Furthermore, peptides corresponding to six partial regions of the amino acid sequence of human TPO shown in SEQ ID NO: 6 (or SEQ ID NO: 7) were synthesized and then used to prepare rabbit polyclonal anti-TPO peptidases. The resulting antibodies were confirmed to recognize human TPO. (Example 45).

It is possible to purify TPO by an affinity column chromatography that uses a column packed with certain molecules that have binding affinity for TPO, such that anti-TPO antibodies, TPO receptors and the like are bound. As a consequence, an anti-TPO antibody column was first prepared by binding the anti-TPO peptide antibody obtained in Example 45. (Example 46).

Purification of human TPO expressed in COS 1 cells using anti-TPO peptide antibody column, and

molecular weight and biological characteristics thereof

Using a culture supernatant from transfection of COS 1 cells with expression vector pHTP1 as a starting material, partially purified TPO was obtained and applied to the anti-TPO antibody column. Since the TPO activity was found in the adsorbed fraction, this fraction was further subjected to a reverse phase column chromatography to purify the protein and examine its molecular weight and biological characteristics.

(Example 47).

Confirmation of activity of partially purified

sample of human TPO expressed in COS 1 cells

TPO-active fraction purified in Example 36, namely TPO sample purified by the step from Capcell Pak CI 300A column from a culture supernatant obtained by transfection of COS 1 cells with expression vector pHTP1, was investigated for its biological. activities to determine whether it exerts a platelet-enhancing effect in the living body. (Example 48).

A crude TPO fraction obtained by a cation exchange column from 33 liters of a culture supernatant prepared by transfected COS 1 cells with expression vector pHTP1 was examined for its biological activities to find that it increased the platelet count in the body.

(Example 49).

Expression of human TPO chromosomal DNA in CHO

cells and confirmation of activity

A vector pDEF202-ghTPO was constructed for use in expression of the human TPO chromosome in CHO cells. (Example 50).

When this vector was introduced into CHO cells, a transforming cell was obtained in which the human TPO chromosomal DNA coding expression vector was integrated into the chromosome. When this cell clone was cultured, the TPO activity was detected in the culture supernatant. (Example 51).

Partial cleaning and activity confirmation of

mutation type human TPO expressed in E. coli

Mutant-type human TPO derived from the human TPO nucleotide sequence-encoding clone, pCFM536/h6T( 1-163 ) and expressed in E. coli was subjected to refolding using guanidine hydrochloride and glutathione to find that the h6T(1-163) thus obtained exerts a platelet count increasing effect in the living body. (Example 53). Mutant type human TPO derived from human TPO nucleotide sequence coding clone, pCFM536/h6T( 1-163) and expressed in E. coli was subjected to refolding using sodium N-aluroyl sarcosinate and copper sulfate and then to a cation exchange chromatography to find that the thus purified h6T(1-163) increased the platelet count in the living body. (Example 54).

In addition, refolding and purification of mutation type human TPO, h6t (1-163) was carried out to apply other procedures. (Examples 60 and 61).

(V) Expression of human TPO cDNA in insect cells

and identification of TPO activity

A recombinant virus for expression of human TPO in insect cells was prepared (Example 58) and expressed in insect cell Sf21, and then the TPO activity was identified in the culture supernatant. (Example 59).

(W) Expression of human TPO (amino acid positions 1 to 163) in CHO cells and purification thereof.

The vector pDEF202-hTPO163 for expression in CHO cells of human TPO protein (referred to as "hTP0163") having amino acids 1 to 163 of the human TPO amino acid sequence shown in SEQ ID NO: 7 was constructed. The expression vector of pDEF202-hTPO163 was transfected into CHO cells and the thus obtained hTP0163-producing CHO cell line was cultivated on a large scale. From the culture supernatant, hTPO 163 was purified. (Example 62 to 65).

(X) Preparation of substitution derivatives of human TPO

A derivative (referred to as "N3/TPO") in which Arg-25 and Glu-231 of human TPO have been substituted by Asn and Lys, respectively, was prepared as well as a derivative (referred to as "09/TPO") in which His-33 has been replaced with Thr.

A plasmid encoding each derivative was transfected into COS7 cells which were then cultured. The TPO activity was detected in the supernatant of the culture. (Example 67).

Derivatives in which one amino acid of h6T(1-163) has been substituted with another amino acid were prepared using the E. coli expression system. The TPO activity was found in each derivative. (Example 94).

(Y) Preparation of insertion or deletion derivatives of human TPO

An amino acid is inserted into or deleted from hTPO 163 to produce its insertion or deletion derivatives, and then COS7 cells were transfected with plasmids encoding the resulting derivatives. In the supernatant of COS7 cell cultures, TPO activity was detected. (Example 68).

A method for measuring TPO activity (an in vitro analysis system) which was used in the present invention is described below as a "reference example".

(Z) Preparation and purification of anti-human TPO antibodies Polyclonal antibodies were prepared using peptide fragments and human TPO (Example 69) and subsequent utilization in Western blot analysis (Example 70) and in the construction of anti-TPO antibody affinity columns. (example 719.

(AA) In vivo activity of human TPO

Purified human TPO was administered to mice with induced thrombocytopenia and changes in platelet count were recorded versus a control group. (Example 72), to (Example 79). The pharmaceutical composition is described as potentially useful in the treatment of thrombocytopenia. (Example 80) to (Example 89).

(BB) TPO activity as a function of MLP receptor binding An analysis is provided where TPO activity is defined on the basis of specific binding interaction with the Mpl receptor. (Example 90) to (Example 93).

Reference example

A. Rat megakaryocyte progenitor cell analysis

(rat CFU-MK assay) system (liquid culture system) Megakaryocytes incorporate and store serotonin in cytoplasmic dense granules through an active, energy-dependent process (Fedorko, Lab. Invest., vol. 36, p. 310-320, 1977). This phenomenon can be observed already in small and mononuclear acetylcholinesterase-positive cells which are considered to be located between CFU-MK and recognizable megakaryocytes (Bricker and Zuckerman, Ecp. Hematol., vol. 12, p. 672, 1984). The amount of serotonin incorporated increases in response to increases in the size of megakaryocytes (Schick and Weinstein, J. Lab. Clin. Med., vol. 98, pp. 607-615, 1981). In addition, it is known that such phenomena are only specific for megakaryocyte cells among bone marrow cells (Schinck and Weinstein, J. Lab. Clin. Med., vol. 98, pp. 607-615, (1981). In the present assay system, highly enriched rat CFU-MK cells (GpHb/ina<+> CFU-MK fraction which will be described later) cultured in the presence of samples to be analyzed, and incorporation of <l4>C-serotonin (<l4>C-hydroxytryptarnine creatine sulfate; <* 4>C-5HT) into megakaryocytes cultured from CFU-MK is measured.

Advantages of this analysis system are that the indirect influence of contaminated cells (for example formation of Meg-CSF activity by contaminated cells stimulated by a certain substance different from the factor of interest, or formation of a certain factor by contaminated cells, which undergo a combined effect with the relevant factor) may be reduced, because the percentage of CFU-MK cells in the GpIIbÆQa"<1>" CFU-MK fraction is markedly high (see the following (analysis method) while the number of contaminating cells is low. In addition, suitable culture conditions can be maintained for a relatively long period, because the total number of cells seeded per well is small. Another advantage is that the number of large-sized mature megakaryocytes cultured from CFU-MK in the presence of an active sample during a culture period can be visually detected during a phase contrast microscope, thus making it possible to qualitatively assess the presence and degree of activity. The results of this qualitative assessment the flour corresponds well to the results of quantitative determination based on the incorporation of <14>c-serotonin. The reliability of the quantitative determination can be further improved by combined application of qualitative assessment.

Analysis method:

Highly enriched rat CFU-MK cells (Gpnb/DIa+ CFU-MK fraction) used in the analysis were prepared by slightly modifying the method of Miyazaki et al. (Exp. Hematol., vol. 20, pp. 855-861, 1992).

Briefly, femur and tibia are removed from Wister rats (male, 8 to 12 weeks old) to prepare bone marrow cell suspension as described. A suspension medium (a solution consisting of 13.6 mM trisodium citrate, 11.1 mM glucose, 1 mM adenosine, 1 mM theophylline, 10 mM HEPES (pH 7.25), 0.5% bovine serum albumin (referred to as "BSA" hereafter) (Path-O -cyte 4; Seikagaku Kogyo Col., Ltd.) and Ca<2+> and Mg<2+_ >free Hank's balanced salt solution: hereafter referred to as "HATCH solution") which is a slightly modified version of megakaryocyte separation media reported by Levin and Fedoroko, Blood, vol. 50, pp. 713-725, 1977) is used for the production of bone marrow cell suspension. The thus prepared bone marrow cell suspension is layered over a Percoll discontinuous density gradient solution (density; 1,050 g/ml (1,063 g/ml/1,082 g/ml)) which is prepared by diluting percoll solution (manufactured by Pharmacia) with HATCH solution, and centrifuged at 20°C at 400 x g for 20 minutes. After centrifugation, the cells between 1,063 g/ml and 1,082 g/ml density layers are recovered. After washing, recovered cells are suspended in Iscove's modified Dulbecco's culture medium (hereafter referred to as sopm "IMDM culture medium") containing 10 % fetal calf serum (hereafter referred to as "FCS"), placed in 100 mm tissue culture plastic dishes and cultured at 37°C for 1 hour in a 5% CO2 incubator. After incubation, non-attached cells were recovered and cultured again in plastic dishes at 37 °C for 1 hour.Unattached cells were suspended in HATCH solution and incubated for 1 hour in 100 mm bacteriological petri dishes to which a mouse monoclonal anti-rabbit plate GpIIb/ma antis toff, P55 antibody (Miyazaki et al., Thromb. Res., vol. 59, pp. 941-953, 1990) had previously been adsorbed. After incubation, non-attached cells were removed by washing through with HATCH solution, and remaining cells adsorbed to immobilized P55 antibody are released by pipetting and collected. In general, 3-4 x 10<5 >cells are obtained from a rat. The cell fraction thus obtained contains highly enriched rat CFU-MK (hereafter referred to as "GpIIb/IIIa<+> CFU-MK fraction"), and it is known that this cell fraction contains about 5 to 10% CFU-MK based on measurements of a colony assay system in the presence of a saturating concentration of rat IL-3. A hybridoma capable of producing the aforementioned P55 antibody has been deposited under deposit number FERM BP-4563 in the National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry on February 14, 1994.

Then, cells of the thus obtained GpIIb/lIIa<+> CFU-MK fraction were suspended in IMDM culture medium containing 10% FCS and dispensed in portions of 10<4> cells per well of a 96 well tissue culture plate. Each well is further supplemented with a standard sample (which will be described in detail in the following), or a sample that was to be analyzed, in order to adjust the final medium volume to 200 (xl/well. The thus prepared plate is placed in a CO2 incubator and incubated for 4 days at 37° C. On the fourth day of incubation, 0.1 [iCi (3.7 KBq) <l>^C-serotonin was added to each well 3 hours before completion of cultivation, and the final incubation is continued at 37° C. After 3 hours of incubation, the plate is centrifuged at 1000 rpm for 3 minutes and the resulting supernatant fluid is removed by aspiration. A 200 µl portion of PBS containing 0.05% EDTA is added to each well, and the plate is centrifuged to wash the plate by removing the resulting supernatant fluid . This washing step is repeated once more. A 200 µl aliquot of 2% Triton X-100 is added to the thus obtained pellet of cells in each well, and the plate is shaken on a plate shaker for about 5 to 10 minutes to thoroughly lyse the cells. A 150 µl portion of the cell lysate thus obtained was transferred into a commercial cover cell-shaped solid state scintillator (Ready Cap; manufactured by Bechman) which is then left overnight at 50°C in an oven to dry the lysate. The next day, the Ready Cap is placed in a glass to measure radioactivity of <14>C by a liquid scintillation counter.

In this case, almost the same results as those from the ^C-serotonin incorporation assay can be obtained when acetylcholinesterase activity in the aforementioned cell lysate is measured according to the method of Ishibashi and Burstein (Blood, vol. 67, pp. 1512-1514, 1986 ).

Standard tests:

First, blood plasma from thrombocytopenic rats was prepared in the following manner for use in the preparation of standard samples.

Normal male Wistar rats (7 to 8 weeks old) were rendered thrombocytopenic by intravenous administration of the aforementioned P55 antibody at a dose of 0.5 mg per animal, twice with an interval of about 24 hours. About 24 hours after the second administration, the blood was collected from the abdominal aorta under ether anesthesia using a 10 ml capacity syringe in which 1 ml of 3.8% (v/v) trisodium citrate solution had been included as an anti-coagulant. The blood sample thus collected was dispensed into plastic centrifuge tubes and centrifuged at 1,200 x g for 10 minutes to recover the blood plasma fraction. The blood plasma fraction thus recovered was re-centrifuged at 1,200 x g for 10 minutes, and the resulting blood plasma fraction was recovered, being careful to avoid pellet consisting of cells, platelets, and the like, and collected. (Blood plasma obtained in such a manner is hereafter referred to as "TRP"). Ca-treated TRP (standard sample C), WGA-agarose column active fraction (standard sample W) or phenyl sepharose 6 FF/LS column active fraction (standard sample P) obtained from TRP in accordance with the procedure of Example 1 were extensively dialyzed against a sufficient volume of IMDM culture medium and used as analytical standards.

In the rat TPO purification process as described in Example 1, standard sample C was used in the early stage, but was changed to standard sample W halfway thereafter and then to standard sample P. Specific activity of standard sample C was tentatively defined as 1, and relative activities of the standard samples W and P were calculated based on the definition. Relative activity of each sample to be analyzed was determined by comparing the dose-response curve of the standard sample to that of the samples to be analyzed, and relative activity of the sample to be analyzed was defined as n where its activity is n times higher than that of standard sample C.

B. Colonyalvsesvstem

In this assay system, bone marrow cells were cultured in a semi-solid culture medium in the presence of a sample to be analyzed, and Meg-SCF activity was measured by counting the number of megakaryocyte colonies formed by proliferation and differentiation of CFU-MK.

Analysis method:

(a) In the case of using unseparated rat bone

marrow cells and the like

A 1 ml final volume of IMDM culture medium containing unseparated rat bone marrow cells, cells obtained in each separation step from the Gpllb/Illa<+> CFU-MK of the aforementioned assay system A, or cells from the GpIIb/IIIa<+> CFU-MK fraction, and 19 \0% FSC, 2 mM glutamine, 1 mM sodium pyruvate, 50 µm 2-mercaptoethanol and 0.3% agar (AGAR NOBLE), manufactured by DIFCO) were placed in a 35-mm tissue culture plastic dish, and after solidification at room temperature, cultured at 37°C in a CO2 incubator. In general, the number of cells per dish is adjusted to 2-4 x IO<5> in the case of unseparated bone marrow cells, 2-5 x IO<4> in the case of cells from the Percoll density gradient step or from adherent washout, or 0.5-2 x 10<3> in the case of cells of the GpIIb/ma<+> CFU-MK fraction. After 6 to 7 days of cultivation, the agar discs were detached from the dishes and placed on glass plates (76 mm x 52 mm). To dry each agar slice, a piece of 50-μm nylon mesh and filter paper was placed, in that order, over the gel. The agar plates thus dried were heat-fixed for 5 minutes at 50°C on a hot plate and dipped for 2 to 4 hours in an acetylcholinesterase staining solution prepared according to the method of Jackson (Blood, vol.42, pp. 413-421 , 1973). When adequate staining of megakaryocytes is confirmed, the agar plates are washed with water, dried, subjected to post-staining with Harris hematoxylin solution for 30 seconds, washed with water and then air dried. Megakaryocyte colonies are defined as three or more tightly clustered acetylcholinesterase-positive cells. (b) In the case of using unseparated mouse bone marrow cells, the colony analysis can be carried out in the same way as procedure (a) above by using 2-4 x 10<5> cells per dish.

(c) In that case using human bone marrow cells

or human marrow cells

Human bone marrow cells or human marrow blood cells can be used as they are, or as CFU-MK fractions enriched in the following way.

First, bone marrow fluid or marrow blood was overlaid on lymphoprep (manufactured by Daiichi Kagaku Col., Ltd) and centrifuged, and the resulting interface leukocyte fraction was recovered. From the cell fraction thus recovered, cells to which biotinylated monoclonal antibodies specific for human surface antigens bind (CD2, CD1 lc and CD 19) were removed with avidin-bound magnetic beads. Cells removed by the magnetic bead method are mainly B cells, T cells, macrophages and some granulocytes. The remaining cells are stained with FlTC-labeled anti-CD34 antibody and PE-labeled anti-HLA-DR antibody, and then a CD34- and HLA-DR-positive fraction is recovered using a cell sorter (for example ELITE manufactured by COULTER ). CFU-MK cells were concentrated in this fraction (which is referred to as the "CD34<+>DR<+> CFU-MK fraction"). Colony analysis of human cells can be carried out in the same way as in the aforementioned case using rat bone marrow cells, with the exception that 3-5 x IO<3> cells of the CD34<+>DR<+> CFU-MK fraction were used in a dish and a mixture of 12.5% human AB plasma and 12.5% FCS was used instead of 10% FCS. The culture period for the formation of megakaryocyte colonies is 12 to 14 days. For the detection of human megakaryocytes, immunological fargin of megakaryocytes was carried out by an alkaline phosphatase-anti-alkaline phosphatase antibody method with a mouse monoclonal antibody specific for a megakaryocyte surface antigen GpHb/IJIa (for example , Teramura et al., Exp. Hematol., vol. 16, pp. 843-848, 1988), and colonies each consisting of three or more megakaryocytes were counted. C. Assay system with a human megacroblast cell line (M-07e assay) M-07e cells, a human megakaryoblast cell line, proliferate in response to GM-CSF, IL-3, SCF, IL-2 and the like (Avanzi et al., J. Cell. PhysioL, vol. 145 , pp. 458-464, 1990). Since these cells also responded to TPO, they could be used as a substitute analysis system for the rat CFU-MK analysis system.

Analysis method:

M-07e cells were maintained in the presence of GM-CSF are recovered, washed thoroughly and then suspended in JMDM culture medium containing 10% FCS. Resulting M-07e cell suspension is dispensed in portions of 10<4> cells into wells of a 96-well tissue culture plate, and each well is further supplemented with a standard sample or a sample to be analyzed, thereby adjusting the final volume to 200 µl/well . The thus produced plate is placed in a 5% CO2 incubator and incubated for 3 days at 37°C. After 3 days of cultivation, 1 (xCi (37 KBq) <3>H-thymidine was added to each well 4 hours before completion of cultivation. After completion of cultivation, the cells are collected on a glass fiber using a cell harvester to measure <3>H- radioactivity with a liquid scintillation counter (eg Beta Plate manufactured by Pharmacia).

An assay system (Ba/F3 assay) in which a mouse pro-B cell line is used

Mouse pro-B cell line Ba/F3 is known as a cell line that proliferates in response to IL3, IL4 and the like (Palacios et al., Cell. vol. 41, pp. 727-734). Since its substrain BF-TE22 has the ability of cell proliferation in response to not only 1L3 and JL4 but also TPO, it can be used in an assay system as a substitute for rat CFU-MK assay or human megakaryoplast cell line-assisted assay (M-07e assay ) as follows. Briefly, BF-TE22 cells growing in Iscove's modified DMEM medium (IMDM:GIBCO) in the presence of 1 ng/ml mouse IL-3 are recovered, washed three times with IMDM and suspended in IMDM medium containing 10% FCS. The cell suspension is then dispensed into wells of a 96-well wood culture plate at a density of 1 x 10<4> cells per well, each well is further supplemented with a standard TPO solution or a sample to be examined and adjusted to a final volume of 200 fil/ well. The resulting plate incubated for 2 to 3 days in a 5% CO2 incubator. On the second or third day, 1 (iCi (37 KBq) <3>H-thymidine was added to each well and the culture continued for an additional 4 hours. Then the cells were collected on a glass fiber filter using a cell harvester to measure <3>H radioactivity with a liquid scintillation counter. In this assay system, almost all cells cultured in media lacking TPO activity died and therefore did not incorporate <3>H-thymidine. However, cells cultured in media containing TPO activity showed active proliferation in a TPO concentration-dependent manner and incorporated <3>H-thymidine In addition, results of this assay parallel those from the M-07e and CFU-MK assays.

The following examples are provided to further illustrate the present invention.

EXAMPLE 1-1

Purification of rat TPO from blood plasma of thrombocytopenic rats induced by anti-platelet antibody administration. Preparation of blood plasma from thrombocytopenic rats induced by anti-platelet antibody administration TRP was prepared as the purification source from approximately 1000 rats in accordance with the procedure described in the aforementioned (reference example) A, rat megakaryocyte precursor cell (CFU-MK) analysis system.

Purification of rat TPO from TRP

To begin with, the purification was carried out by using TRP from about 1000 rats on the assumption that the TPO content in TRP would be 1 per three million for the most part, and slightly less than 1 pmol of partially purified rat TPO was obtained. Although difficult in general, an attempt was made to analyze its partial amino acid sequences and three partial amino<y>resequences were obtained, but with low accuracy. These sequences coincided with or are close to the amino acid sequence of a serine protease inhibitor (SP1) produced in liver. It was unclear from these results whether TPO has a structure corresponding to that of serine protease inhibitor or whether the sequence is derived from those of contaminating proteins other than TPO. Using these uncertain sequences, cloning of the TPO gene from a rat cDNA library was carried out, but potential TPO-encoding genes could not be obtained. Intensive studies were then carried out based on the assumption that such a failure is due to low purity and an insufficient quantity of analyzed sample. In order to obtain a purified final sample necessary for amino acid analysis, purification was carried out in accordance with a procedure described below (Example 1-2). It was surprisingly estimated from the results in (Example 1-2) that TPO content in TRP or XRP (will be described below) was an extremely small amount as one hundred million to one billion of total plasma proteins, so that complete purification of extremely difficult.

EXAMPLE 1-2

Purification of rat TPO from blood plasma of thrombocytopenic rats induced by total X-ray or gamma irradiation (preparation of blood plasma of thrombocytopenic rats induced by total X-ray or gamma irradiation)

Normal male Wister rats (7 to 8 weeks old) were rendered thrombocytopenic by whole-body irradiation with X-rays or gamma rays at a sublethal dose (6 to 7 Gy). Blood was collected from the rats on the 14th day after irradiation and a blood plasma fraction (hereafter referred to as "XRP") was prepared in the same manner as the aforementioned method for the preparation of TRP.

In this case, XRP (about 8L) was prepared from a total of about 1,100 rats as the purification source.

Purification of TPO from XRP

Blood plasma from about 1,100 rats was too large for and was processed simultaneously because its total protein content reached 493,000 mg. As a consequence, the blood plasma sample was divided into 8 lots, each corresponding to approximately 100 rats, in the following purification steps (1) to (4). In subsequent purification steps (5) to (7), the sample was divided into six batches. In and after purification step (8), a crudely purified sample from the total blood plasma of approximately 1,100 rats was used. In the following, typical examples of each cleaning step are described in the case of a lot (XW9) and a batch (XB6).

In all the purification steps, the TPO activity was measured by using the rat CFU-MK analysis system as described in (reference example). If not stated otherwise, all the purification steps were carried out at 4_C, except reverse phase chromatographies and Superdex 75pg filtration in the presence of surfactant which were carried out at room temperature. The protein analysis was carried out by a Coomassie stained binding assay (a reagent system manufactured by PIERCE, catalog no. 23236X) or a method using bicinchoninic acid (a reagent system manufactured by PIERCE, catalog no. 23225).

An overview of the purification is shown in table 1.

(1) Calcium chloride treatment, centrifugation and

<p>rotase inhibitor treatment of rat blood plasma

In the case of Lot No. XW9

XRP (742 ml; protein concentration, 54.8 mg/ml; total protein, 40,686 mg) corresponding to approximately 100 rats, which has been stored at -80°C, was thawed and dispensed into polypropylene centrifuge tubes (manufactured by Nalgene). Calcium chloride powder was added to each tube to a final concentration of 100 mM. After overnight incubation at 4°C, the resulting mixture was centrifuged at 8000 rpm for 60 minutes. To the resulting TPO-active supernatant fluid (742 ml; protein concentration, 54.9 mg/ml; total protein, 40,740 mg) was added a protease inhibitor p-APMSF ((p-aminodiphenyl)methanesulfonyl fluoride hydrochloride, manufactured by Wako Pure Chemical Industries, Ltd.). , catalog no. 010-10393) to a final concentration of 1 mM. The sample thus obtained was used in the following Sephadex G-25 column after the buffer exchange step.

In this way, calcium chloride/p-APMSF treatment of whole XRP derived from about 1,100 rats (total volume, 8,184 ml; total protein, 493,000 mg) was carried out by dividing it into 11 lots, each lot being equivalent to about 100 rats, and carrying out processing these lots one by one. The resulting 11 samples were processed separately in subsequent Sephadex G-25 column steps.

(2) Sephadex G-25 (buffer extraction)

in that case of Lot No. XW9

Supernatant fluid obtained after calcium treatment in step (1) (742 mL; protein concentration, 54.9 mg/mL; total protein, 40,740 mg) was applied at a flow rate of 40 to 70 mL/min onto a Sephadex G-25 column (manufactured by Pharmacia Biotech , catalog no. 17-0033-03; diameter, 11.3 cm; height 47 cm) which had previously been equilibrated with 20 mM Tris-HCl, pH 8. Elution was carried out with the same buffer. A 1300 ml portion of the eluate before elution of protein was discarded. When UV absorption was detected in the eluates, the fractions were pooled until the electrical conductivity reached 500 iiS/cm, thus recovering the TPO-active protein fraction which had been exchanged with 20 mM Tris-HCl, pH 8 (1.377 ml; protein concentration, 27.56 mg/ ml; total protein, 37.882 mg; protein yield 93%). Relative activity of TPO in the fraction was found to be 2.3.

As a result of Sephadex G-25 column treatment of all lot samples, a total of 21,117 ml of Sephadex G-25 column TPO active fraction was obtained (total protein 408,300 mg; average relative activity 2; total activity 864,600).

(3) O-Sepharose FF (strong anion liquid chromatography) in it

the case of Lot No. XW9

The TPO-active fraction obtained in step above (2) by Sepharose G-25 treatment (1.377 ml; protein concentration, 27.5 mg/ml; total protein, 37.841 mg; relative activity, 2.3) was applied to a Q-Sepharose FF ( manufactured by Pharmacia Biotech, catalog no. 17-0510-01; diameter, 5 cm; height, 27 cm) at a flow rate of 40 ml/min, eluted with 20 mM Tris-HCl (pH 8) and a fraction Fl passed through the column was collected (3.949 mL; protein concentration, 0.98 mg/mL; total protein, 3.870 mg; relative activity, 0).

Then the buffer was changed to 20 mM Tris-HCl (pH 8) containing 175 mM NaCl to elute a TPO-active fraction F2 (4.375 ml; protein concentration, 5.36 mg/ml).

Finally, a fraction F3 (1.221 ml; protein concentration, 3.9 mg/ml; total protein, 4.783 mg; relative activity, 3.8) was eluted with 20 mM Tris-HCl (pH 8) containing 1000 mM NaCl. Total protein in TPO-active fraction F2 was found to be 23,440 mg, and the protein yield of F2 in this step was 61.9%. In addition, relative activity of TPO was increased to 6.8.

As a result of applying all lots of Sephadex G-25 TPO-active fractions to Q-Sepharose FF, a total of 35,842 ml of Q-Sepharose FF TPO-active fraction F2 (total protein, 314,384 mg; mean relative activity 9; total activity 2,704,000).

(4) Wheat germ a<g>glutinin ( WGA)- agarose

(lectin affinity chromatography)

In the case of Lot No. XW9

TPO-active fraction F2 obtained in the above step (3) by Q- Sepharode FF treatment' was divided into three portions and applied to a WGA-agarose (manufactured by Honen Corp., catalog no. 800273; diameter 5 cm; height 22.5 cm ) at a flow rate of 5 ml/min, and eluted with Dulbecco's isotonic phosphate buffer (DPBS). A fraction F1 that passed through the column was collected (9,336 ml; protein concentration 2.30 mg/ml; total protein 21,407 mg; relative activity 6.9).

Then the buffer was changed to 20 mM sodium phosphate buffer (pH 7.2) containing 0.2 M N-acetyl-D-glucosamine (GlcNAc, prepared by Nacalai tesque, catalog no.005-20), 150 mM NaCl and 0.02% sodium azide, and the resulting eluates was collected and concentrated using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cut-off of 8000; manufactured by Filtron), thereby obtaining a WGA-agarose adsorbing TPO-active fraction F3 (2.993 ml; protein concentration 0.376 mg/ml).

Total protein in TPO-active fraction F2 was found to be 1,125 mg, and the protein yield of F2 in this step was 4.8%. In addition, relative activity of TPO was increased to 101. The F2 fraction thus obtained was stored at -80°C.

As a result of applying all lots of Q-Sepharose FF TPO-active F2 fractions to WGA-agarose, a total of 33,094 ml of WGA-agarose TPO-active fraction F2 (total protein, 15,030 mg; average relative activity, 132; total activity, 1,987,000).

TSK- eel AF- BLUE 650 MH (triazine dye affinity chromatography)

In that case of Rate No. XB6

WGA-agarose adsorbing TPO-active fraction of Lot No. XW8 and WGA-agarose adsorbing TPO-active fraction F2 of Lot No. XW9 obtained in the above step (4) by starting with 215 rat equivalents of XRP were combined as a batch no. .XB6 (5.947 ml, protein concentration, 0.388 mg/ml; total protein, 2.319 mg, relative activity, 150).

A 5.974 mL portion of the combined sample was mixed with 0.85 mol NaCl for 1000 mL (296.76 h total) to make a 6.132 mL solution having a final NaCl concentration of 0.822 M, and the resulting solution was applied at a flow rate of 7 ml/min. to a TSK-gel AF-BLUE 650 MH column (TOSOH CORP., catalog no. 08705; diameter, 5 cm; height, 23 cm) which had been pre-equilibrated with 20 mM sodium phosphate buffer ((pH 7.2) containing 1 M NaCl .

After completion of the application, protein was eluted using 20 mM sodium phosphate buffer (pH 7.2) containing 1 M NaCl at a flow rate of 10 ml/min. The resulting eluates were collected and concentrated using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cut-off 8000), thereby obtaining a passed-through fraction Fl (543 ml; protein concentration, 2.05 mg/ml; total protein, 1.112 mg; relative activity, 31 ).

Then the elution buffer was changed to 2M NaSCN to obtain an eluted TSK gel AF-BLUE 650 MH adsorbing TPO-active fraction F2 (1.427 ml; protein concentration, 0.447 mg/ml).

Total protein in TPO-active fraction F2 was found to be 638 mg, and the protein yield of F2 in this step was 27.5%. In addition, relative activity of TPO was increased to 1500.

As a result of applying all batches of WGA-agarose adsorbing TPO-active F2 fractions to TSK-gel AF-BLUE 650 MH, a total of 10,655 ml of TSK-gel AF-BLUE 650 MH TPO-active fraction F2 was obtained (total protein 4,236 mg; mean relative activity, 905; total activity 3,834,000).

Fenvl Sepharose 6 FF/ LS (hydrophobic interaction chromatography)

In that case of Rate No. XB6

A 1,424 ml portion of TSK-gel AF-BLUE 650 MH TPO-active fraction F2 (1,424 ml; protein concentration, 0.447 mg/ml; total protein, 638 mg; relative activity, 1,500) was mixed with 1.5 mol ammonium sulfate powder per 1,000 ml (282.2 g total) to make a 1.581 mL solution having a final ammonium sulfate concentration of 1.35 M.

The resulting solution was applied at a flow rate of 7 ml/min to a phenyl Sepharose 6 FF (Low Sub) column (Pharmacia Biotech, catalog no. 17-0965-05; diameter, 5 cm; height 10 cm) which was equilibrated in leading edge with 50 mM sodium phosphate buffer (pH 7.2) containing 1.5 M ammonium sulfate. After completion of the application, elution was performed using 36 mM sodium phosphate buffer containing 0.8 M ammonium sulfate at a flow rate of 10 ml/min. The resulting eluates (approximately 3,160 ml) were collected and concentrated using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cut-off 8000), thereby obtaining a fraction Fl (485 ml; protein concentration, 0.194 mg/ml; total protein, 94.2 mg ; relative activity, 0).

Then the elution buffer was changed to 20 mM sodium phosphate buffer (pH 7.2) to obtain an eluted TPO active fraction F2 (about 3,500 ml). The eluted fraction was concentrated using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cut-off 8000) and a sample for analysis was withdrawn. At this step, protein concentration and total protein in TPO-active fraction F2 (220 ml) were found to be 1.45 mg/ml and 319 mg, respectively, and the protein yield of F2 at this step was 50.0%. Relative activity of TPO was found to be 1,230.

As a result of applying all batches of TSK-gel AF-BLUE 650 MH TPO-active F2 fractions to phenyl Sepharose 6 FF/LS, a total of 1,966 ml of phenyl Sepharose 6 FF/LS TPO-active fraction F2 was obtained (total protein 2,762 mg; average relative activity 847; total activity 2,339,000).

(7) Sephacryl S- 200 HR (gel filtration chromatography)

In the case of rate No. XB6 (cfr, fig. 1)

Phenylsepharose 6 FF/LS TPO-active fraction F2 (217 mL; protein concentration, 1.45 mg/mL; total protein, 315 mg; relative activity, 1.230) was mixed with 144.8 mL of 5 M NaCl solution to make a 362 mL solution that has a final NaCl concentration of 2 M, and the resulting solution was concentrated to about 50 ml using an ultrafiltration unit with a YM 3 membrane (76 mm in diameter, manufactured by Amicon Corp.).

To this was added the same volume (50 ml) of 8 M urea, in order to obtain approx. 100 ml of a solution containing 1 M NaCl and 4 M urea as final concentrations. This was concentrated to about 80 ml, made into a sample of 88.78 ml and then applied to a Sephacryl S-200 HR column (manufactured by Pharmacia Biotech, catalog no. 17-0584-01; 7.5 cm in diameter and 100 cm in height) .

The elution was then carried out with DPBS at a flow rate of 3 ml/min, and the eluates after a void volume of 1200 ml were collected in 45 ml portions in 60 polypropylene tubes. Analysis was performed on a second each tube and the remainder stored at -85°C until Sephacryl S-200 HR treatment of all samples derived from 1,100 rats was completed. Based on analysis results, fractions of Batch No. XB6 were grouped as follows.

In this way, all batches of TPO-active F2 fractions obtained by phenylsepharose 6 FF/LS were separately subjected to Sephacryl S-200 HR, analyzed and stored at -85°C. After completion of S-200 HR treatment of all batches and immediately prior to subsequent performance of reverse phase chromatography (YMC-Pack PROTEIN-RP), these stored samples were thawed and concentrated using an ultrafiltration unit with a YM 3 membrane (76 mm in diameter, manufactured by Amicon Corp.) to obtain the following two samples. The concentrated sample of TPO-active fraction F2 obtained by Sephcryl S-200 HR is hereafter referred to as "high molecular weight TPO sample F2" and that from F3 as "low molecular weight TPO sample F3".

High molecular weight TPO sample F2 and low molecular weight TPO sample F3 are pooled fractions of different elution ranges in gel filtration chromatography as appropriate, and the term "high molecular weight" and "low molecular weight" therefore do not mean their actual molecular weights.

Low molecular TPO sample F3 and high molecular TPO sample F2 are separately subjected to subsequent purification steps.

Purification of low molecular weight TPO sample F3 is described in the following steps (8) to (11).

(8) YMC- Pack protein- RP (reverse phase chromatography')

Low molecular weight TPO sample F3 (total protein 50.3 mg; protein concentration 0.184 mg/ml; relative activity 20,000; total activity 1,007,000; total volume, 274 ml) obtained in step (7) above was mixed with a solvent A (0.025% trifluoroacetic acid (TFA)) and a solvent B (1-propanol containing 0.025% TFA) to obtain a solution that has a total volume of 508.63 ml and final concentrations for propanol, TFA and protein of approximately 20%, 0.012% and 0.0989 mg/ ml. Insoluble materials formed during preparation of this solution were separated by centrifugation, and the resulting supernatant was divided into two 254.3 ml (25.2 mg protein) portions and applied at a flow rate of 2 ml/min to a column packed with YMC-Pack PROTEIN-RP (manufactured by YMC, catalog no. A-PRRP-33-03-15; 3 cm in diameter and 7.5 cm in height) which had been previously equilibrated with 30% B. The precipitate resulting from centrifugation was dissolved in 20 ml of 20 mM sodium acetate (pH 5.5) containing 5 mM CHAPS (3-[3-cholamidopropyl)dimethylammonio]-1-propanesulfonate; manufactured by Dojindo Laboratories, catalog no. 75612-03-3) and applied to the column.

After the sample was loaded, about 50 ml of a solvent (solvents A:B = 3:1) was passed through the column to obtain a carry-through fraction. After that, a development program was started (120 minutes of linear gradient from 30% B to 45% B), and the eluates were collected in 36 polypropylene tubes in 10 ml portions. This process was again repeated for the remaining sample by using the same collection tube, so that a total of 36 fraction tubes were obtained each containing 20 ml of eluate. The fraction passed through was directly concentrated to 20 ml using an ultrafiltration unit with a YM 3 membrane (76 mm in diameter, manufactured by Amicon Corp.).

A 0.1 ml portion of each of the fractions passed through and 20 ml fractions of tubes from 1 to 36 was mixed with 20 ml of 5% BSA, dried by evaporation, dissolved in 0.25 ml of IMDM assay culture medium and then analyzed to specify TPO- active faction. As a result, the TPO activity was found in the fractions with tubes no. 17 to 27 (36.0 to 43.0% propanol concentration range) which were combined and used as a low molecular weight TPO sample F3-derived YMC-Pack PROTEIN-RP TPO-active fraction F2. This fraction was stored at -85 °C until use in subsequent YMC-PAck CN-AP steps.

Low molecule TPO sample F3-derived YMC-Pack PROTEIN-RP

TPO-active fraction F2

(9) YMC-PAck CN-AP (reverse phase chromatography')

A 214.9 ml portion of low molecular weight TPO sample F3-derived YMC-PAck PROTEIN-RP TPO activity fraction F2 (total protein, 2.79 mg, protein concentration, 0.0130 mg/ml; relative activity, 130,000; total activity, 36,300) obtained in steps (8) above was mixed with 0.6 ml of 50% glycerol and concentrated to 1.8 ml. The concentrate was finally adjusted to a volume of 5 ml containing 20% or less propanol and about 6% glycerol.

The concentrate was divided into 5 portions for use in the following column operation (0.555 mg protein and 1 ml volume in each operation). Each of the thus divided samples was applied at a flow rate of 0.6 ml/min. to a column packed with YMC-Pack CN-AP (manufactured by YMC, Catalog No. AP-513; 6 mm in diameter and 250 mm height) which has been pre-equilibrated with 15% B, using 0.1% TFA as solvent A and 0.05% TFA-containing 1-propanol as solvent B. After the application, a propanol concentration was increased from 15% B to 25% B, and 65 minutes of linear gradient from 25% B to 50% B was carried out . After completion of the final operation (5th), 1 ml of a solution having the same composition excluding protein was applied to the column and induced in the same manner to recover TPO activity which is one in the column. Since the same polypropylene tube was used to collect the eluates from 6 operations, a total of 44 tubes each containing 7.2 ml of eluate were obtained.

A 30 µl portion of each of the thus obtained fractions (1/240 portion of each fraction) was mixed with 20 µl of 5% BSA, dried and evaporated, dissolved in 0.24 ml of IMDM assay culture medium and then analyzed to specify TPO-active fractions . As a result, strong TPO activity was found in the fractions in tubes No. 28 to 33 (37.0 to 42.0% propanol concentration range) which were combined and used as a low molecular weight TPO sample F3-derived YMC-Pack CN-AP main TPO-active fraction FA .

Low molecular weight TPO sample F3-derived YMC-Pack CN-AP TPO-active fraction FA

Capcell Pak Cl 300 (final reverse phase chromatography)

A 32.12 ml portion of 43.20 ml low molecular weight TPO sample F3-derived TPO-active fraction FA (total protein 0.372 mg; protein concentration, 0.00863 mg/ml; relative activity, 800,000; total activity, 297,500) was obtained in the above step (9) mixed with 0.2 ml 50% glycerol and concentrated to obtain 0.1 ml glycerol solution.

This solution was mixed with 2 ml of a solution of solvent A (0.1% TFA); solvent B (1-propanol containing 0.05% TFA) = 85:15 (15% B) to prepare 2.1 ml sample containing about 14% propanol, about 4.8% glycerol and 0.177 mg/ml protein. The sample thus prepared was applied to a column packed with Capcell Pak Cl 300A (manufactured by Shiseido Col., Ltd, catalog no. Cl TYPE.SG300A; 4.6 mm in diameter and 250 mm in height) which had been previously equilibrated with 15% B and developed for 65 min with a linear gradient from 27% B to 38% B at a flow rate of 0.4 ml/min. The eluates were collected in 72 polypropylene tubes in 0.6 ml portions.

A 3 µl portion of each of the thus obtained fractions (1/200 portion of each fraction) was mixed with 20 µl of 5% BSA, the medium was exchanged with 225 µl of IMDM analysis culture medium and the thus 75-fold diluted fraction was analyzed.

A 1 µl portion of each fraction (1/600 fraction) was sampled for use in electrophoresis, dried and evaporated and treated at 95°C for 5 minutes with 10% of a non-reducing sample buffer for SDS-PAGE. The sample thus treated was subjected to SDS-PAGE using a 15-25% SDS-polyacrylamide precast gel (manufactured by Daiichi Pure Chemicals Col., Ltd.) and then stained with a silver stain kit of 2D-Silver StainH "DAIICHI" (manufactured by Daiichi Pure Chemicals Co., Ltd., Catalog No. 167997; and hereinafter referred to as "silver color kit"). As for molecular weight markers, [DAnCHIj-III and low molecular weight markers (manufactured by Daiichi Pure Chemicals Co., Ltd., Catalog No. 181061; hereinafter referred to as "DPCDI") were used.

As a result of the analysis above, significant DPO activity was found in the fractions in tubes no. 35 to 43 (30.0 to 32.5% propanol concentration range). Of these, fractions from tubes no. 36 to 42 (30.5 to 32.5% propanol concentration range) were combined and used as the main TPO-active fraction FA. The results are shown in fig. 2.

When the protein content deduced from the chromatogram was compared with the analysis results, the fraction thus obtained was found to have a total protein of 39.6 Hg, a protein concentration of 9.4 µg/ml, a relative activity of 4,890,000 and a total activity of 193,600. Examination of the SDS-PAGE pattern of the TPO-active fraction numbers 36 to 42 showed the presence of a band whose staining density correlates with strength of activity. In addition, the molecular weight of this unreduced band was found to be 17,000 to 19,000 when compared to that of standard molecular weight proteins on the same gel, which had been reduced, thus showing that this band is a strong candidate for TPO.

(11) Extraction of TPO-active protein from electrophoresis gel

(15% SDS-PAGE)

Example of analysis of TPO-active fraction FA

Of 4,200 ul low molecular weight TPO sample F3-derived TPO-active fraction FA (total protein 39.6 ug; protein concentration 9.4 ug/ml; relative activity 4,890,000; total activity 193,600) obtained in the above step (10), 5.5 ul (1/ 764 fraction) for use in active protein extraction and 2.5 µl (1/1680 fraction) for use in silver staining transferred to respective sample tubes, dried and evaporated, reacted with 10 µl of a non-reducing sample buffer for SDS-PAGE at 37°C for 1 hour and then had the opportunity to stand for a further 18 hours at room temperature, in order to carry out the SDS reaction.

Pre-stained low range marker (manufactured by Bio-Rad Laboratories, Inc., 161-0305) and the aforementioned DPCin were used as molecular weight markers. Using a microslab gel, 15% SDS-PAGE of these samples was carried out at 4°C in accordance with the commonly used procedure (Laemmli, Nature, vol. 227, pp. 680-685, 1970). After completion of electrophoresis, gel portions to be subjected to silver staining were immediately cut out with a knife, placed in a fixative solution, and then stained using the aforementioned silver staining kit.

Separately from this, all molecular weight ranges of the gel to be used for detection of activity were cut with a knife into 34 slices, and each gel slice having a width of 1.5 to 2.5 mm was crushed with a slightly modified method of Kobayashi (Kobayashi , M., Seikagaku (Biochemistry), vol. 59, no. 9, 1987, published by the Japanese Biochemical Society). Each gel slice thus crushed into smaller fragments was mixed with 0.3 ml of an extraction buffer (20 mM Tris-HCl, pH 8,500 mM NaCl, 0.05% BSA) and shaken for 6 hours at 4°C to carry out extraction.

To this was added 500 mM potassium phosphate (pH 6.8) to a final concentration of 20 mM. After 1 hour of shaking at 4°C, the resulting mixture was transferred to an ultra-free C3GV 0.22 nm filtration unit (manufactured by Millipore Corp., UFC3 OGV OS) and centrifuged at 1,000 x g (4,000 rpm) for 15 minutes to remove precipitated SDS and for remaining resulting supernatant fluid. The filtrate was transferred to an ultrafree C3-LGC ultrafiltration unit (nominal molecular weight cut-off of 10,000, manufactured by Millipore Corp., UFC3 LGC 00) and centrifuged at 3000 x g (7000 rpm). When the volume of the concentrated solution reached about 50 µl, 300 µl of 20 mM sodium phosphate buffer (pH 7.2) was added and again subjected to ultrafiltration.

Ultrafiltration using 300 µl of 20 mM sodium phosphate buffer was repeated twice to remove residual SDS. Then, a similar step was repeated to change to assay culture medium to prepare a sample (300 µl) which was then sterilized and subjected to TPO activity measurement.

Three protein bands were clearly detected by silver staining and showed apparent molecular weights of about 17,000 to 19,000, 14,000 and 11,000 based on DPCDI molecular weight marker.

In the preceding step (10), a band was observed which has a correlation between activity strength and color density and which has an apparent molecular weight of about 17,000 to 19,000 in electrophoresis of the Capcell Pak Cl column TPO-active fraction. In the experiment in this step (11), the TPO activity was also found in the band which has an apparent molecular weight of about 17,000 to 19,000 (cf. fig. 3).

On the basis of the results above, it was confirmed by TPO-active protein was purified in the active fraction of the Capcell Pak Cl 300A column to a detectable level on the electrophoresis gel. Based on silver-stained density of the band obtained by 15% SDS-PAGE, the amount of the relevant TPO protein (an apparent molecular weight of about 17,000 to 19,000) in the total TPO-active fraction was determined to be about 1.7 µg.

Purification of high molecular weight TPO sample F2 was described in the following steps (12) to (15).

YMC- Pack PROTEIN- RP ( reverse phase chromatography)

High molecular weight TPO sample F2 (total protein, 257 mg; protein concentration 0.894 mg/ml; relative activity, 7,840; total activity, 2,015,000; total volume, 287 ml) obtained in the above step (7) was mixed with 95.8 ml (1/ 3 volume of the sample) of a solvent B (1-propanol containing 0.025% TFA) to obtain a solution that has a total volume of 383 ml and final concentrations of propanol, TFA and protein of approximately 25%, 0.06% and 0.621 respectively mg/ml. A solution of 0.025% TFA was used for solvent A. Insoluble materials formed during the preparation of input sample were separated by centrifugation, and the resulting supernatant fluid was divided into six 62.3 ml portions (42.8 mg protein) and applied at a flow rate of 2 ml /min to a column packed with YMC-Pack PROTEIN-RP (manufactured by YMC, catalog no. A-PRRP-33-03-15; 3 cm in diameter and 7.5 cm in height) which had been pre-equilibrated with 30% B The precipitate resulting from centrifugation was dissolved in 10 ml of 20 mM sodium acetate (pH 5.5) containing 5 mM CHAPS also applied to the column.

After the sample was loaded, about 50 ml of a solvent (solvent A:B = 3:1) was passed through the column to obtain a carry-through fraction. Then a development program was started (120 min of linear gradient from 30% B to 45% B) and the eluate was collected in 24 polypropylene tubes in 15 ml portions. This process was repeated for six divided samples by using collection tubes, so that a total of 24 fraction tubes were obtained each containing 90 ml of eluate. The pass-through fraction and tube #1 fraction were combined and directly concentrated to 90 mL using an ultrafiltration unit with a YM3 membrane (76 mm in diameter, manufactured by Amicon Corp.).

A 0.3 ml portion of each of the completed fractions plus red No. 1 fraction and

the fractions from tubes #2 to 24 were mixed with 10 µl 5% BSA, dried by evaporation, dissolved in 0.30 ml IMDM assay culture medium and then analyzed to specify TPO active fractions. As a result, TPO activity was found in the fractions in tubes no. 10 to 15 (34.0 to 39.5% propanol concentration range) which were combined and used as a high molecular weight TPO sample F2-derived YMC-Pack PROTEIN-RP TPO-active fraction F2. This fraction was stored at -85°C until use in subsequent YMC-PAck CN-AP steps.

High molecular weight TPO sample F2-derived YMC-Pack PROTEIN-RP-TPO-active fraction F2

(13) Superdex 75 pg (gel filtration chromatography in the presence of CHAPS)

A 538.2 ml portion of high molecular weight TPO sample F2-derived YMC-Pack PROTEIN-RP TPO active fraction F2 (total protein 11.3 mg; protein concentration 0.021 mg/ml; relative activity 227,000; total activity 2,565,000) obtained in step (12) above was mixed with 0.6 ml of 50% glycerol and concentrated by evaporation. To this was added 18 ml of 20 mM CHAPS, followed by stirring and then incubated for 1 hour at 4°C. Next, the first sample was applied to a column packed with HiLoad 26/60 Superdex 75 pg (manufactured by Pharmacia Biotech, catalog no. 17-1070-01; 2.6 cm in diameter and 60 cm in height) and eluted with DPBS containing 5 mM CHAPS at a flow rate of 1 ml/min. A 4 ml portion of each sample (protein concentration 0.466 mg/ml; protein 1.86 mg) was applied to the column.

In this case, the Superdex 75 column operation was repeated 6 times by counting the entire YMC-Pack PROTEIN-RP TPO-active fraction in 6 portions. The eluates of each operation were collected in 5 ml portions in 45 tubes, so that 45 fractions each containing 30 ml of eluate were finally obtained after completion of 6 column operations.

A 0.1 ml portion of each of the fractions thus obtained was mixed with 10 µl of 5% BSA, dried by evaporation, dissolved in 0.25 ml of IMDM assay culture medium and then analyzed to specify TPO-active fractions. As a result, TPO activity was found in the fractions with tubes No. 13 to 31 (molecular weight range from 78,000 to 3,000) which were combined and used as a high molecular weight TPO sample F2-derived Superdex 75 pg TPO active fraction F2.

High molecular TPO sample F2-derived Superdex 75 pg TPO-active fraction F2

YMC- Pack CN- AP (reverse phase chromatography)

A 513.2 ml portion of high molecular weight TPO sample F2-derived Superdex 75 pg TPO-active fraction F2 540 ml (molecular weight, 78,000 - 3,000) (total protein, 1.11 mg; protein concentration 0.00216 mg/ml; relative activity, 1,750,000; total activity 1,943,000) obtained in step (13) above was mixed with 1 hour volume of a solvent B (1-propanol containing 0.05% TFA) and applied at a flow rate of 0.6 ml/min to a column packed with YMC -Pack CN-AP (manufactured by YMC, catalog no. AP-513; 6 mm in diameter and 250 mm height) which had been equilibrated beforehand with 15% B using a solvent A (0.1% TFA) and solvent B. After application, the propanol concentration was increased from 15% B to 25% B, and a 65 minute linear gradient from 25% B to 50% B was carried out.

Before starting the YMC-Pack CN-AP column chromatography, a 1/20 portion of the total loaded sample was subjected to a sample operation to confirm the correct recovery of activity. Then the remaining 19/20 portion was divided into two samples to carry out the operation twice. In other words, the column operation was repeated three times. Since the same 44 polypropylene tubes were used to collect the eluates from three operations, a total of 44 tubes each containing 3.6 ml of eluate were obtained.

A 5 µl portion of each of the fractions thus obtained (1/720 portion of each fraction) was mixed with 0.25 ml of IMDM assay culture medium and then analyzed to specify TPO-active fractions. As a result, a markedly strong TPO activity was found in the fractions in tubes No. 24 to 30 (36.0 to 42.0% propanol concentration range) which were combined and used as a high molecular weight TPO sample F2-derived YMC-Pack CN-AP main TPO active fraction FA.

High molecular weight TPO sample F2-derived YMC-Pack CN-AP TPO active

fraction FA

(15) Capcell Påk Cl 300 A (final reverse phase chromatography)

A 24.66 ml portion of 25.20 ml high molecular weight TPO sample F2-derived YMC-Pack CN-AP TPO active fraction FA (total protein 0.606 mg; protein concentration, 0.0246 mg/ml; relative activity, 700,000; total activity 424,000) obtained in step ( 14) above was mixed with 0.4 ml of 50% glycerol and concentrated by evaporation.

In this way, 2 ml of a concentrated sample were obtained with a propanol concentration of a few percent, a glycerol concentration of 10% and a protein concentration of 0.303 mg/ml. This was applied to a column packed with Capcell Pak Cl 300A (manufactured by Shiseido Co., Ltd., catalog no. Cl TYPE:SG300A; 4.6 mm diameter and 250 mm height) which was pre-equilibrated with 15% B by applying a solvent A (0.1% TFA) and solvent B (1-propanol containing 0.05% TFA), and developed for 65 min by linear gradient from 27% B to 38% B at a flow rate of 0.4 ml/min. The eluates were collected in 72 polypropylene tubes in 0.6 µl portions.

A 0.75 ni portion of each of the thus obtained fractions (1/800 portion of each fraction) was mixed with 20 ul 5% BSA, the medium was changed to 225 ul IMDM analysis culture medium and the thus 300 times diluted fraction was analyzed.

A 2 µl portion of each fraction (1/300 fraction) was sampled for use in electrophoresis, dried by evaporation and treated at 95°C for 5 minutes with 10 µl of a non-reducing sample buffer for SDS-PAGE. The thus treated sample was subjected to SDS-PAGE using a 15-25% SDS-polyacrylamide precast gel (manufactured by Daiichi Pure Chemicals Co., Ltd.) and then stained using the above silver staining kit. DPCDI described above was used as molecular weight markers.

As a result of the above analysis, significant TPO activity was found in the fractions in tubes numbered 33 to 39 (29.5 to 31.5% propanol concentration range). Of these, fractions of tubes no. 34 to 39 (30.0 to 31.5% propanol concentration range) were combined and used as the main TPO-active fraction FA. Examination of the SDS-PAGE pattern of the main TPO-active fraction showed the presence of a band whose staining density correlates with activity strength, with a molecular weight in the range from 17,000 to 19,000 under non-reducing conditions, and this corresponded to the situation with low-molecular TPO sample F3-derived TPO - active fraction described in the above-mentioned step (10).

Example 2

Analysis of partial amino acid sequences of purified rat TPO

Amino acid sequence analysis of rat TPO protein in Capcell Pak Cl 300A column TPO-active fraction FA obtained in purification step (10) in Example 1 was carried out in accordance with the procedure of Iwamatsu (Iwamatsu et al., "Shin Kiso Seikagaku Jikken-ho (New Basic Biochemical Experiments)", vol. 4, pp. 33-84, pub. Maruzen; Iwamatsu, A. Seikagaku (Biochemistry), vol. 63, no. 2, pp. 139-143, 1991; Iwamatsu, A., Electrophoresis, vol. 13, pp. 142-147, 1992). This means that the sample was subjected to SDS-PAGE and transferred electrically onto a polyvinylidene difluoride (PVDF) membrane. After reduction and S-alkylation of the protein on the PVDF membrane, the thus treated protein was cleaved into peptide fragments by systematic and stepwise restrictive enzymatic hydrolysis in situ with three proteases, and resulting peptide fragments at each cleavage step were separated and purified by reverse phase chromatography and analyzed for their amino acid sequences by a high-sensitivity amino acid sequence determination method. The following describes the process in detail.

Example of analysis of TPO relevant protein in low molecular weight TPO sample F3-derived Capcell Pak Cl 300A TPO fraction FA

(1) Concentration of Capcell Pak Cl 300A column

TPO fraction (room number 36 to 42)

Of the 4200 µl low molecular weight TPO sample F3-derived Capcell Pak Cl 300A column TPO-active fraction FA (tube numbers 36 to 42) obtained in purification step (10) in Example 1 (total protein 39.6 µg; protein concentration, 9.4 µg/ml; relative activity 4,890,000; total activity 193,600), 4,151 µl (98.8% of total fraction) was used in amino acid sequence analysis. Total protein deduced from the chromatogram was 39.1 µg, and of this approximately 1.6 µg was the TPO protein stained with silver after SDS-PAGE and which has a molecular weight of approximately 17,000 to 19,000.

This sample was mixed with glycerol and concentrated by evaporation to obtain 5 µl of a glycerol solution. To this was added a non-reducing buffer for SDS-PAGE and 1 M Tris-HCl (pH 8) to adjust the pH, thus obtaining about 25 µl of the sample containing 200 mM Tris-HCl (pH 8.0), 50 mM Tris-HCl (pH 6.8), 1.1% SDS; 2 mM EDTA, 0.02% bromophenol blue and 30% glycerol.

The thus prepared sample was kept at room temperature for 14 hours and then treated at 60°C for 5 minutes to carry out the proper SDS reaction without excess heating.

(2) Electrophoresis

Micro-slab gels (4.0% acrylamide stacking gel and 15% acrylamide separation gel) were prepared and SDS-PAGE was carried out at room temperature for 2 hours at a constant current of 12.5 mA and then at 17.5 mA. Pre-stained Low Range Marker (Bio-Rad, 161-0305) and DOCDI were used as molecular weight markers. Immediately after electrophoresis, the resulting protein molecules are transferred onto a PVDF membrane (see following steps).

A portion of the sample was also used in another electrophoresis using 15-25% polyacrylamide precast gel (Multi Gel 15/25 manufactured by Daiichi Pure Chemicals Co., Ltd., catalog no. 211072) under non-reducing conditions or after reduction of the sample with dithiothreitol (DTT). When the resulting gel was stained using a silver staining kit as described above, it was confirmed that the molecular weight of the protein band expected to be TPO was about 19,000 under a reducing condition and the purity of the TPO protein in the Capcell Pak Cl 300A column was a few percent. In addition, since the mobility of the band varied under non-reducing and reducing conditions, it was suggested that the protein in question contains at least one disulfide bond.

Transfer of Protein to PVDF Membrane by Electroblotting and Band Detection Protein transfer to a PVDF membrane (ProBlott, manufactured by Applied Biosystems, catalog no. 400994) was carried out for 1 hour at a constant current of 160 mA (11 to 17 V) using a semi-dry transfer apparatus (Model KS-8460, manufactured by Marysol). A solution consisting of 0.3 M Tris and 20% methanol (pH 10.4) was used as the anolyte, a solution consisting of 25 mM Tris and 20% methanol (pH 10.4) as the transfer membrane solution and a solution consisting of 25 mM Tris, 40 mM aminocaproic acid and 20% methanol (pH 10.4) as cathlyte.

When the thus transferred membrane was stained with a Ponceau S staining solution (0.1 g Ponceau S and 1 ml acetic acid in 100 ml water), several bands were detected and one of these bands was confirmed to be that of the protein in question and had a molecular weight of about 19,000. This band was excised for use in subsequent peptide fragment generation steps.

(4) Peptide fragment formation. peptide mapping and

amino acid sequence analysis

In order to carry out systematic fragmentation of the TPO candidate protein that has been transferred and S-alkylated after reduction on the PVDF membrane, stepwise restrictive enzymatic hydrolysis was carried out using the following three proteases.

First cleavage: lysyl endopeptidase (Achromobacter lyticus m497-l, manufactured by Wako Pure Chemical Industries, Ltd., Catalog No. 129-02541)

second cleavage: endoproteinase Asp-N (manufactured by Boehringer-Mannheim Corp., catalog no. 1054 589)

third cleavage: trypsin-TPCK (manufactured by Worthington Biochemical, catalog no. 3740).

Peptide fragments obtained in each enzyme cleavage are recovered, applied to a Wakosil-II5C18 C18 reverse-phase column (manufactured by Wako Pure Chemical Industries, Ltd., 2.0 mm in diameter and 150 mm in length) and eluted by using a solvent A

(0.05% TFA) and a solvent B (isopropanol:acetonitrile = 7:3,0.02% TFA), 30 minutes of a linear gradient from 1% B to 50% B at a flow rate of 0.25 ml/min and at a column temperature of 30°C. In this way, the peptide maps were created (see fig. 4). Resulting peptide fragments were recovered and subjected to Edman degradation

by using a gas phase amino acid sequencer (PPSQ-2, manufactured by Shimadzu Corp.). Then, each amino acid recovered from the sequencer, whose N-terminus had been coupled with PTH, was identified using C18 reverse phase column chromatography by isocratic elution. The results are summarized below.

Amino acid sequences of peptide fragments obtained by first cleavage with lysyl endopeptidase

Amino acid sequences of peptide fragments obtained by second cleavage with endopeptidase Asp-N

Amino acid sequences of peptide fragments obtained by third digestion with trypsin-TPCK

In the sequences above, those shown in parentheses are the N-terminal and amino acid residues that can be deduced from systematic enzyme digestion.

(5) Analysis of homology of obtained amino acid sequences

(homology search)

In order to know whether the thus obtained amino acid sequences are contained in a known reported protein or whether it is a protein that has similar sequences, they were analyzed using sequence analysis software Mac Vector (Kodak International Biotechnologies, Inc.)- Regarding information about known proteins or known genes, the Entrez Release 6 database was used (National Center for Biotechnology Information, National Library of Medicine, National Institute of Health, USA, published 15 August 1993). Databases included in this are as follows.

Entrez Release 6 database

NCBI-GenBank, 15 August 1993, (Release 78.0)

EMBL, 15 July 1993 (Release 35.0 plus updates)

DDBJ, 15 July 1993

SWISS-PROT, April 1993 (Release 25.0)

PIR, 30 June 1993 (Release 37.0)

PDB, April 1993

PRF, May 1993

dbEST, 15 July 1993 (Release 1.10)

U.S. and European patents.

As a result, the sequence (K)DSFLADVK of AP12 coincided completely with an internal sequence KDSFLADVK of a rat corticosteroid-binding globulin (CBG) precursor (PIR database accession no. A40066; Smith and Hammond; "Rat corticosteroid-binding globulin: primary structure and messenger ribonucleic acid levels in the liver under different physiological conditions". Mol. Endocrionol., (1989), 3,420-426).

Further detailed investigation showed that a sequence KQYYESE (SEQ ID NO: 193) which has a high similarity to the sequence (K)XYYESZ (X is A, S, G, M or Q and Z is E or K) of AP3 is contained in rat CBG amino acid sequence. In rat CBG, sequences corresponding to those in AP 12 and AP3 are linked together and thus form an internal amino acid sequence KDSFLADVKQYYESE (SEQ ID NO: 190).

With regard to the amino acid sequences of fragments other than AP 12 and AP3, no protein or gene was found whose similarity could be taken into account.

(Example 3)

Analysis of biological characteristics of TPO derived from blood plasma of thrombocvtooenic rats

(1) In the rat CFU-MK assay system (liquid culture system)

As a typical example, Figure 5 shows a dose response curve for the use of a TPO sample partially purified from thrombocytopenic rat plasma (TPO-active fraction F2 from YMC Pack-Protein-RP column described in purification step (8) in example 1-2). When the cultured cells were observed periodically under a microscope, the generation and maturation of megakaryocytes, namely an increase in cell number, was observed, probably together with an increase in the number of megakaryocytes. As a particularly significant change, process formation of megakaryocytes was found on the fourth day which was the last day of culture (they were hardly observable on the third day). Such process formation is called cytoplasmic process formation (Leven and Yee, Blood, vol. 69, pp. 1046-1052, 1987) or proplatelet process formation (Topp et al., Blood, vol. 76, pp. 912-924, 1990), which is thought to be a platelet precursor structure further differentiated from mature megakaryocytes and is thought to be the final step of megakaryocyte differentiation observed so far in vitro. As such a morphological change was observed with high frequency when using the TPO sample alone, it is possible that this factor alone can stimulate the proliferation and differentiation of CFU-MK, can produce mature megakaryocytes and can finally release platelets.

(2) In the colony analysis system

When a TPO sample partially purified from blood plasma of thrombocytopenic rats was examined with the colony assay system using unseparated rat bone marrow cells, cells from each of the separation/concentration steps or a GpIIb/IIIa<+> CFU-MK fraction, of the rat plasma-derived TPO stimulated the formation of a megakaryocyte colony. Compared to the megakaryocyte colonies induced by other cytokines such as rat IL-3, mouse GM-CSF or human EPO, the TPO-induced megakaryocyte colonies are characterized by the fact that each colony consists of a smaller number of megakaryocytes, but each megakaryocyte is larger in size, meaning greater maturity. In addition to the fact that TPO produced no or only few colonies in the other cell lines, the Meg-CSF activity of TPO can be considered megakaryocyte-specific. On the basis of these facts, it is clear that TPO has different biological properties from those of other cytokines such as rat IL-3, mouse GM-CSF, and human EPO and exhibits unique Meg-CSF activity.

The TPO sample partially purified from blood plasma of thrombocytopenic rats also exerted its Meg-CSF activity on CD34<+>, DR<+> cell fractions derived from human bone marrow cells or human spinal cord blood cells and induced significant numbers of human megakaryocyte colonies, indicating that this factor had no type specificity .

(Example 4)

Specialization of rat TPO-producing cells

(1) Screening of rat TPO-producing organs Screening and specialization of rat TPO-producing organs was performed to secure an mRNA source for use in cloning a rat TPO gene. First, bone marrow, lungs, liver and spleen were taken periodically from rats rendered thrombocytopenic by P55 antibody administration, their cells (organ sections for lungs and liver) were cultured and activities in the resulting culture supernatants were examined using the CFU-MK assay system. However, this initial attempt did not yield any distinctive results. Accordingly, further attempts were made to culture hematocytes prepared by collagenase perfusion in the liver of thrombocytopenic rats induced by P55 antibody administration, taking into account a report indicating a relationship between the liver and TPO production in rats (Siemensma et al., J. Lab. Clin. Med., vol. 86, pp. 817-833, 1975). When the resulting supernatant was applied to a WGA-agarose column and then the adsorbed fraction was fractionated on the eb Vydac phenyl reverse phase column, substantially similar activity was found at the same position of the rat plasma derived TPO activity in the rat CFU-MK assay system. Weak but the same activity was also found in the culture supernatant of normal rat hepatocytes. These results strongly suggested that the liver will be one of the TPO producing organs. (2) Screening of rat TPO-producing cell lines On the basis of the above results, screening of rat TPO-producing cell lines was performed. First, each of 20 rat liver-derived cell lines were cultured in respective subculture media until the cells reached near-confluent level, the culture media was exchanged with the same respective media supplemented with 5% FCS, and culture was continued for 3 days. When each of the resulting culture supernatants was partially purified according to the procedure described in step (1) above to investigate the existence of TPO production, TPO activity was clearly found in three rat hepatic parenchymal cell-derived cell lines, namely McA-RH8994 cells (ATCC accession number CRL1602; Becker et al., "Oncodevelopmental Gene Expression" edited by Fisman and Seli, Academic Press, NY. pp. 259-270, 1976; obtained from Dainippon Pharmaceutical Co., Ltd.), H4-H-E cells (ATCC accession number CRL1548; Pitot et al. al., Nat. Cancer Inst. Monogr., vol. 13, pp. 229-245, 1964; obtained from Dainippon

Pharmaceutical Co., Ltd.) and HTC cells (Thompson et al., Proe. Nati. Acad. Sei. USA, vol. 56, pp. 296-303, 1966; obtained from Dainippon Pharmaceutical Co., Ltd.).

(3) Detailed analysis of TPO activity in McA-RH8994 cells.

H4-II-E cells and HTC cells

Biochemical and biological properties of the TPO-active proteins secreted from these three rat cell lines were investigated in detail and compared with those in rat plasma derived

TPO.

McA-RH8994 cells were suspended in alpha-MEM(-) liquid culture medium containing 10% FCS and transferred to 175 cm 2 plastic tissue culture flasks at 1 x 10<6> cells/bottle. After 3 days of cultivation at 37°C in a 5% CO2 incubator, the medium was exchanged with IMDM culture medium containing 5% FCS and cultivation was continued for another 3 days to recover the resulting culture supernatant. H4-H-E cells were suspended in Dulbecco's modified Eagle's liquid culture medium (glucose 4.5 g/l, hereafter referred to as "DMEM") containing 10% FCS transferred to 175 cm<2> plastic tissue culture flask at 5 x 10<5> cells/bottle. After 3 days of cultivation at 37°C in a 5% CO2 incubator, the medium was exchanged with IMDM culture medium containing 5% FCS and cultivation was continued for another 3 days to recover the resulting culture supernatant. HTC cells were also suspended in DMEM liquid culture medium containing 5% FCS and transferred into a 175 cm<2> plastic tissue culture flask with 2 x 10^ cells/bottle. After 3 days of cultivation at 37°C in a 5% CO2 incubator, the medium was exchanged with IMDM culture medium containing 5% FCS and cultivation was continued for another 3 days to recover the resulting culture supernatant.

Using a 2 liter portion of the culture supernatant thus produced from each of the three cell lines, partial purification of the cell line-derived TPO was carried out according to the procedure described for the purification of TPO from XRP described in examples 1 to 2. An outline of the results is described below.

First, each of the culture supernatants was concentrated about 6 times using an ultrafiltration apparatus and the medium was exchanged with 20 mM Tris-HCl buffer (pH 8.0) using a Sephadex G-25 column. The resulting eluate was applied to a Q-Sepharose FF column, the column was washed with 20 mM Tris-HCl (pH 8.0) and the adsorbed fraction was eluted with 20 mM Tris-HCl (pH 8.0) containing 175 mM NaCl. These thus eluted fractions were put on a WGA-agarose column, the column was washed with PBS and the adsorbed fraction was eluted with 20 mM sodium phosphate (pH 7.2) containing 0.2 M GlcNAc and 0.15 M NaCl. Resulting eluates were applied to a TSK-gel AF-BLUE 650MH column, the column was washed with 20 mM sodium phosphate (pH 7.2) containing 1 M NaCl and the adsorbed fraction was eluted with 2 M NaSCN. The resulting eluate was applied to a phenyl sepharose 6 FF/LS column, the column was washed with 50 mM sodium phosphate (pH 7.2) containing 1.5 M ammonium sulfate followed by 0.8 M ammonium sulfate containing 36 mM sodium phosphate and then the adsorbed fraction was eluted with 20 mM sodium phosphate (pH 7.2). The adsorption fraction thus obtained was concentrated and then fractionated using a reverse phase Vydac protein C4 column (manufactured by The Separations Group, catalog no. 214TP51015; 1 cm in diameter and 15 cm in height. That is, the samples were placed on the C4 column that had been equilibrated beforehand with 20% B and then the elution was carried out for 90 minutes by linear gradient from 20% B to 40% B at a flow rate of 1 ml/min using 0.1% TFA in developing solvent A and a propanol containing 0.05 % TFA in development solvent B. As a result, each of the TPO active proteins derived from these cell lines was eluted with a 1-propanol concentration of 30 to 43%.

When the TPO activities in the samples in each purification step were measured using the rat CFU-MK analysis system, the results showed that the TPO activity on the three cell lines behaved with a similar pattern to that of the XRP-derived TPO (see examples 1-2). Relative activity, activity yield and the like at each purification step measured by the rat CFU MK analysis system are shown in table 2.1 addition, when the TPO activity in the eluates from the final reverse phase column was measured by the rat CFU MK analysis system, the TPO activities of the three cell lines and XRP-derived TPO found in the same peak area. When column analyzes TPO active fractions from the reverse phase column were performed using non-adherent cells obtained in the adherence dilution step on the separation of the concentrations of rat GpIIb/nia<+> CFU-MK, the elution patterns of Meg-CSF activity were close to those of TPO activity measured by the rat CFU-MK assay system (see Table 3). In addition, treated similarly to the file trap by XRP-derived TPO, each of the TPO activities of three cell lines no or few colonies of the other cell lines.

Each of the collected active fractions from the reverse phase column was subjected to SDS-PAGE according to the procedure described in examples 1-2. When protein is extracted from the resulting gel and TPO activity was measured by the rat CFU-MK assay system, an apparent molecular weight of XRP-derived TPO, McA-RH8994 cell-derived TPO, and H4-II-E cell-derived TPO was found to be 17,000, respectively. to 22,000, 33,000 to 39,000 and 31,000 to 38,000 and the HTC cell-derived TPO showed apparent molecular weights of 17,000 to 22,000 and 28,000 to 35,000.

These results thus revealed that the TPO proteins produced by McA-RH8994 cells, H4-U-E cells and HTC cells are equivalent to the XRP-derived TPOs in terms of their biological properties, although their biochemical properties are slightly different from each other in terms of apparent molecular weight. A notable finding of the present invention is the possibility that a TPO-active protein molecule contained in blood may be a product of partial digestion at its specific or irregular site in its producing cells or after its secretion from the producing cells. The possible existence of TPO genes (mRNA and cDNA) with different lengths has also been suggested.

(Example 5)

Construction of expression vector (pEF18S) for cDNA library A vector into which the cDNA fragment can be integrated easily and which can exhibit a high expression efficiency of the cloned cDNA was constructed for use in cloning and expression of TPO cDNA. That is, an expression vector pEF18S was constructed from the expression vector pME18S by replacing the SRoc promoter with an elongation factor la (EFla) promoter known as a high expression promoter (see Figure 6). The promoter of elongation factor 1a was obtained by partial digestion of 1 µg of an expression vector pEF-BOS (Mizushima et al., Nucleic Acids Res., vol. 18, p. 5322, 1990) with restriction enzymes HindHI and EcoRI, exposing the result to a 2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a cDNA fragment of about 1200 bp and purification of the DNA fragment using a Prep-A gene DNA purification kit (manufactured by Bio-Rad Laboratories, Inc.; a kit for use in selective purification of DNA as performed by porous silica matrix-supported DNA adsorption, developed on the basis of the procedure reported by Willis et al. in Bio Techniques, vol.9, pp. 92-99, 1990).A 100 ng portion of the cDNA fragment thus obtained was ligated with 50 ng of an expression vector pME18S (Liu et al., Proe. Nati. Acad. Sei. USA, vol. 90, pp. 8957-8961, 1993) which has been previously digested with Hindm and EcoRI Cells of a commercial host strain, Competent High E. coli DH5 (manufactured by Toyobo Co., Ltd., a comp etent E. coli strain prepared by a modified method according to Hanahan et al., J. Mol. Biol., vol. 166, pp. 557-580, 1983), was transformed with the thus constructed vector. After culturing the transformed cells, 12 colonies were randomly selected and plasmid DNA was purified from each colony. A total of 10 colonies containing the relevant plasmids (pEF18S) were obtained by analyzing the restriction enzyme according to the pattern for these DNA molecules. Then a clone selected from there was cultured to produce a large amount of plasmid DNA.

Purification of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). That is, the clone pEF18S prepared above was cultured overnight at 37°C in 50 ml of LB medium (1% Baco-tryptone, 0.5% Bacto-yeast extract, 0.5% NaCl) containing 50 (in g/ml ampicillin and the resulting cells collected by centrifugation was suspended in 4 ml of TEG-lysozyme solution (25 mM Tris-HCl, pH 8, 10 mM EDTA, 50 mM glucose, 0.5% lysoxym). To this was added 8 ml of 0.2 N NaOH/1% SDS solution and then 6 ml of 3M potassium/5 M acetate solution to thoroughly suspend the cells. After centrifugation of the suspension, the resulting supernatant fluid was treated with phenol-chloroform (1:1), mixed with the same volume of isopropanol, and then centrifuged. The resulting pellet was dissolved in TE solution (10 mM Tris-HCl, pH 7.5, 1 mM EDTA) and treated with RNase and then with phenol-chloroform (1:1), followed by ethanol precipitation. The resulting pellet was again dissolved in TE solution to which it then NaCl and polyethylene glycol 3000 were added to a final concentration of respectively 0.63 M and 7.5%. After centrifugation, the pellet was dissolved in TE solution precipitated with ethanol. In this way, about 300 µg of plasmid DNA was obtained. A 100 µg portion of DNA was completely cut with restriction enzymes EcoRI and Noti and subjected to 0.8% agarose gel (manufactured by FMC BioProducts) electrophoresis and the thus isolated vector fragment was purified using of the Prep-A-Gene DNA purification kit (manufactured by Bio-Rad laboratories, Inc.) to obtain 55 µg of plasmid DNA which was used in the following cDNA library construction.

(Example 6)

Purification of mRNA from McA-RH8994 cells

McA-RH8994 cells which showed relatively high activity in the colony analysis in example 4 were selected as material for use in rat TPO cDNA cloning and subjected to the following experiments.

Isolation of total RNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). That is, McA-RH8994 cells were grown to confluence in 15 culture dishes of 90 mm diameter. After removal of the liquid medium, cells in each dish were thoroughly suspended in 0.8 ml of a 5 M guanidine solution (5 M guanidine thiocyanate, 5 mM sodium citrate (pH 7.0), 0.1 Mfi-mercaptoethanol, 0.5% sodium sarcosyl sulfate), and the resulting suspension in all the dishes were collected in a tube and the total volume was adjusted to 20 ml with the same guanidine solution. The resulting mixture, which became viscous due to cell destruction, was subjected to 20 repetitions of suction/exhalation with a 20 ml syringe fitted with an 18G needle and then a 21 g needle until the mixture became almost non-viscous. An 18 ml portion of 15.7 M CsCl-0.1 M EDTA (pH 7.5) was used as a pad in a polyallomer centrifuge tube for use in a Beckman SW28 rotor and the above mixture about 20 µl was carefully placed over the pad in such a way that the layers was not destroyed and the tube was almost filled. The tube thus prepared was subjected to 20 hours of centrifugation at 25,000 revolutions per minute at 20° C. The resulting pellet was washed 2 times with a small amount of 80% ethanol, dissolved in TE solution, extracted with phenol/chloroform (1:1) and then subjected to ethanol precipitation with a 1/10 volume of 3 M sodium acetate and 2.5 volumes of ethanol.In this way, about 2.5 mg of total RNA was obtained from about 108 cells.

Purification of poly (A)+ RNA from total RNA was performed using OligotexTM-dT30 (Super) (manufactured by Japan Synthetic Rubber/Nippon Roche; oligo dT molecules are immobilized on latex particles by covalent binding and purification of poly (A)+ RNA can be preserved similarly to the use of an oligo dT column which is generally used for such purposes.As a result, 20 ng of poly(A)+ RNA was obtained from about 500 ng of total RNA.

(Example 7)

Construction of rat cDNA library

Double-stranded cDNA with an EcoRI recognition site at its 5'-tagged end and a Noti recognition site at its 3'-end was synthesized from 5 ng of poly (A)+ RNA prepared in Example 6 using the Time SaverTM cDNA synthesis kit (manufactured by Pharmacia ; a cDNA synthesis kit based on the modified method of Okayama and Berg, Mol. Cell. Biol., vol. 2, pp. 161-170, 1982) and DIRECTIONAL CLONING TOOLBOX (manufactured by Pharmacia; a kit of a primer 5-AACTGGAAGAATTCGCGGCCG -CAGGAA(T)18-3' (SEQ ID NO: 15) containing a Noti recognition site for use in the synthesis of cDNA and adapter 5-AATTCGGCACGAG-3' (SEQ ID NO: 16) and 5-CTCGTGCCG-3' (SEQ ID NO: 17) for use in the addition of an EcoRI recognition sequence). The cDNA thus prepared was ligated with 1.2 ng of the expression vector pEF18S (see Example 5) which had been cut beforehand with EcoRI and Noti and transformed into 8.4 ml of the above Competent High E. coli DH5 (manufactured by Toyobo Co., Ltd .). As a result, 5.3 x 10 5 transformants were obtained.

(Example 8)

Production (cloning) of rat TPO cDNA fragment by PCR

The 530,000 clones of McA-RH8994 cDNA library constructed in Example 7 were grown overnight in 50 ml LB medium containing 50 ng/ml ampicillin and McA-RH8994 cDNA library plasmids were purified from the resulting cells collected by centrifugation, using QIAGEN-tiplOO (manufactured by DIAGEN; an anion exchange silica column for DNA purification). About 200 ng of plasmid DNA was obtained.

Two anti-sense nucleotide primers AP8-1R and AP8-2R corresponding to the amino acid sequences of the peptide fragment AP8 described in Example 2 and two sense nucleotide primers EPloc-1 and EFla-2 corresponding to the first intron of the human elongation factor la of the plasmid vector pEF-18S (see fig.6) used for the preparation of the cDNA library, was synthesized. Synthesis of these primers was carried out using a 394DNA/RNA synthesizer (manufactured by Applied Biosystems; a synthesizer based on the B-cyanoethylamidite method) their purification was carried out using an OPC column for synthetic DNA purification (manufactured by Applied Biosystems; a reverse phase silica gel column for use in the purification of synthetic DNA with trityl groups). Each of the thus synthesized and purified DNA molecules was dissolved in TE solution to a concentration of 50 µM and stored at -20°C until use. The synthetic mononucleotide used in the following procedures was synthesized and purified in the same way.

Primers AP8-1R and AP8-2R are mixed primers comprising 17 consecutive nucleotides into which deoxyinosine is incorporated as reported by Takahashi et al.

(Proe. Nati. Acad. Sei. USA, vol. 82, pp. 1931-1935, 1985).

Primers EFla-1 and EFla-2 consisting of 21 and 20 nucleotides, respectively, were synthesized based on the nucleotide sequence and parts 1491 to 1512 and 1513 to 1532 of the genomic sequence reported by Uetsuki et al. (J. Biol. Chem. vol. 264, pp. 5791-5798

(1989).

Using 3 µg of the McA-RH8994 cDNA library plasmid as a template, AP8-1R (500 pmol) and EFla-1 (100 pmol) as primers and Gene Amp™ PCR Reagent Kit with Amplitaq™ DNA polymerase (manufactured by Takara Shuzo Co ., Ltd.; a kit of a thermo-stable taql polymerase, a reaction buffer and dNTPs for use in PCR), PCR was performed by GeneAmp™ PCR system 9600 (manufactured by Perkin-Elmer; a thermal apparatus for PCR) (100 ni volume; heated to 95°C for 2 min, a total of 35 cycles, each cycle comprising denaturation at 95°C for 1 min, annealing at 40°C for 1 min and synthesis at 72°C for 1 min and final incubation at 72°C for 7 minutes). To improve the specificity of the thus amplified DNA fragments, PCR would be performed (100 µl volume; heating to 95°C for 2 minutes, a total of 35 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 45° C for 1 minute and synthesis at 72°C for 1 minute and final incubation at 72°C for 7 minutes) using 1 µl of the thus prepared PCR reaction solution as template, and EFla-2 (100 pmol) and AP8- 2R (500 pmol) as primers.

The reaction solution thus prepared was subjected to 2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a DNA fragment of about 330 bp as the main product of PCR which was then purified using the above Prep-A-Gene DNA purification kit (manufactured by Bio-Rad Laboratories, Inc.). Using T4 DNA ligase (manufactured by Life Technologies), the thus purified DNA fragment was subcloned into the pCRTMU vector (a vector for the use of TA cloning of PCR products, manufactured by Invitrogen). As the thermostable polymerase for use in PCR had a terminal transferase activity the amplified DNA fragment was directly subcloned into the pCRTMU vector which has a 5'-dT cohesive end by means of the enzyme's ability to add a deoxyadenylic acid to the 3' end of it PCR-amplified DNA. Plasmid DNA molecules were purified from 28 clones selected randomly from the resulting clones using QIAGEN-tipl00 (manufactured by DIAGEN) and their nucleotide sequence was determined by a 373A DNA sequence (a fluorescence sequence, manufactured by Applied Biosystems), using a Taq Dye DeoxtTM Terminator Cycle Sequencing Kit (manufactured by Applied Biosystems; a kit for use in PCR-assisted nucleotide sequence determination using fluorescent dyes, based on the dideoxy method of Sanger et al., Proe. Nati. Acad. Sei. USA, vol. 74 , pp. 5463-5467, 1977).

When a DNA fragment encoding three amino acid residues (UeAThr/Ser)-Val-Pro, of the amino acid sequence from AP8 shown above is adjacent to the AP8-2R primer, was selected from the thus prepared DNA fragments and its entire length was sequenced, it contained cDNA with 261 bp. Positions 173 to 175 of this cDNA fragment coded for ethionine and the coding frame matched the amino acid frame of AP8. When the sequence in position 173-175 in the DNA fragment matched the sequence according to Kozak (Kozak, M., Cell. vol. 44, p. 238-292, 1986) it appeared that this cDNA fragment contains a translation initiation region that codes for the N-terminus to rat TPO protein. This cDNA fragment was given the name Al. The nucleotide sequence of the Al fragment excluding the vector sequence and an amniotic acid sequence derived therefrom, shown in the attached sequence listing (SEQ ID NO: 1).

(Example 9)

Screening of rat TPO cDNA cloned by PCR

The above cDNA library was divided into pools of about 10,000 clones, grown overnight in 1 ml of the above LB medium containing 50 (Ag/ml ampicillin and then subjected to plasmid DNA extraction using the automated plasmid isolation apparatus Pl-100 (VER -3.0, manufactured by Karabo Industries, Ltd.; an automated plasmid DNA extraction apparatus based on a modified procedure of the alkali SDS method described in Molecular Cloning, Sambrook et al., Cold Spring Harbor Laboratory Press, 1989).Separately from this, they were two following oligonucleotides synthesized on the basis of cDNA fragment Al and purified.

5'CGAGGGTGTACCTGGGTCCTG 3" (SEQ ID NO: 31) (a sense sequence of positions 1 to 17 of the sequence according to SEQ ID NO: 1;

CGAG stands for an adapter sequence)

5'CAGAGTTAGTCTTGCGGTGAG 3' (SEQ ID NO: 32) (an anti-sense sequence position 212 to 232 of the sequence according to SEQ ID NO: 1)

Using 1/30 volume of each plasmid DNA sample obtained against the same template, with the synthesized oligonucleotides as primers and GeneAmp™ PCR Reagent Kit with AmpliTaq™ DNA Polymerase (manufactured by Takara Shuzo Co., Ltd), PCR was performed with GeneAmp ™ PCR System 9600 (manufactured by PERKIN-ELMER) (a total of 30 cycles, each cycle consisting of denaturation at 94°C for 30 seconds, annealing at 66°C for 30 seconds and synthesis at 72°C for 1 minute). As a result, a specific band of 236 bp was detected in 3 of the 100 stocks examined. When one of these three pools was subdivided into subpools of about 900 clones to extract plasmid DNA and PCR was performed in the same manner, a specific band was detected in 3 of the 100 subpools tested. When one of these subpools was further divided into pools of 40 clones and the same screening process was performed, the specific band was found in 3 of the 100 pools tested. After the stocks in question were grown on an LB plate (LB medium containing 15% agar) 50 µg/ml ampicillin was added and each of the colonies thus formed was subjected to plasmid DNA extraction and PCR in the same way. As a result, a positive band was detected in 2 of the 100 clones examined.

(Example 10)

Sequencing of rat TPO cDNA

Purification of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). Each of the two clones obtained in example 9 was grown overnight in 50 ml LB medium containing 50 µg/ml ampicillin and about 300 µg plasmid DNA was obtained after purification in the same way as described in example 6.

The plasmid DNA thus obtained was subjected to sequencing in the same manner as described in Example 8 using the Taq Dye DeoxyTM Terminator Cycle Sequencing Kit (manufactured by Applied Biosystems) to determine the complete nucleotide sequence containing the Al fragment. As a result, the nucleotide sequences of the two clones corresponded completely with each other, showing that they are the same clone. One of these clones was selected and the plasmid carried by this clone was named pEF18S-A20. Its nucleotide sequence and an amino acid sequence derived therefrom are shown in the sequence listing (SEQ ID NO: 2).

The sequence shown in the sequence listing (SEQ ID NO: 2) has distinct characteristics. The sequence downstream of the 172 bp 5' untranslated region encodes an amino acid sequence rich in hydrophobic amino acid that is believed to be a protein secretion signal sequence consisting of 21 amino acids starting with methionine. This protein contains 126 amino acid residues and a 3' untranslated sequence of 1025 nucleotides and a poly A tail following the termination codon (TAA). This protein contains a sequence corresponding to the amino acid sequence AP8 analyzed in example 2 (amino acid no. 1 to 12 in SEQ ID NO: 2), but does not contain a consensus sequence for N-glycosylation. The 1624 to 1629 nucleotide sequence located near the terminus of the 3' untranslated sequence is different from the consensus sequence, but appears to be a potential polyadenylation sequence.

The vector pEF18S-A2a carried by an E. coli strain DH5 has been deposited by the present inventors on February 14, 1994 under deposit number FERM BP-4565, in the National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

(Example 11)

Expression of rat TPO cDNA in COS 1 cells and confirmation of TPO activity Transfection of plasmid pEF18S-A2a into COS 1 cells was performed as follows

by slight modification of the DEAE-dextran method which includes chloroquine treatment (Sompayrac et al., Proe. Nati. Acad. Sei. USA, vol. 78, pp. 7575-7578, 1981; Luthman et al., Nucl. Acids Res. , vol. 11, pp. 1295-1308, 1983). COS 1 cells (ATCC CRL1650) were suspended in the above DMEM medium containing 10% FCS, put up in 100

mm plastic plastic tissue culture dishes and cultured at 37°C in a 5% CO 2 incubator until the cells reached about 40% confluency. Separately, plasmid pEF18S-A26 (10 µg) dissolved in 30 µl HBS (21 mM HEPES, 145 mM NaCl, pH 7.1) was added to 4 ml of the DMEM culture medium to which 500 µg/ml DEAE-dextran (prepared by Pharmacia), 80 [mu]m chloroquine (manufactured by Sigma Chemical Co.), 8% (v/v) of HBS and

9% (v/v) of Nu serum (manufactured by Collaborative Research, Inc.). The thus prepared mixture was added to the above cultured COSI cells which had been washed twice with DMEM culture medium and cultivation was continued for 5 hours at 37°C in the 5% CO 2 incubator. Then the culture supernatant in the dish was removed by suction and the remaining cell in the dish was washed twice with DMEM culture medium, added 15 ml of DMEM culture medium containing 10% FCS and cultured at 37°C for 3 ten, 5 days in the 5% C02 incubator to recover the resulting culture supernatant.

The culture supernatant thus obtained was thoroughly dialyzed against IMDM culture medium and evaluated with the rat CFU-MK analysis system. TPO activity was found in a dose-dependent manner in the culture supernatant of the COS 1 cells in which the plasmid pEF18S-A2cc was expressed (fig. 7). Similar to the case of XRP-derived TPO, many megakaryocytes formed elongated cytoplasmic processes on the fourth day of culture. In contrast, TPO activity was not found in the culture supernatant of COS 1 cells into which the plasmid pEF18S alone had been transfected (fig. 7). In the M-07e assay system, M-07e cell proliferation-enhancing activity was also found in a dose-dependent manner in the culture supernatant of COS 1 cells in which the plasmid pEF18S-A2oc was expressed, but not in the culture supernatant of COS 1 cells in which the plasmid pEF18S alone was expressed. These results demonstrate that A2a contains a cDNA that codes for a protein that has a TPO activity.

Next, a partially purified TPO sample was prepared to investigate the ability of this TPO activity to promote platelet production in vivo. First, COS 1 cells transfected with the plasmid pEF18S-A2a were cultured for three days in a serum-free culture medium to which 0.2 ng BSA had been added, whereby 5.8 liters of serum-free culture supernatant were obtained. Protease inhibitor p-APMSF was added to the serum-free culture supernatant to the final concentration of 1 mM and the mixture was filtered using a 0.22 µm filter.b To 5.793 ml of the resulting filtrate (protein concentration, 0.229 mg/ml; total protein, 1.326 mg ; relative activity, 1000; total activity 1,326,000) 0.85 mol of NaCl was added for each 1000 ml (288 g total), to obtain 5849 ml of solution containing 0.822 M NaCl. The solution thus prepared is provided, at a flow rate of 7 ml/min., on a TSK gel AF-BLUE 650 MH column (manufactured by Tosoh Corp., catalog no. 08705; 5 cm diameter and 6 cm height) which had been pre-equilibrated with 20 mM sodium phosphate (pH 7.2) containing 1 M NaCl. After completion of the sample application, approximately 7,900 ml of eluate was passed through the column, eluted with 20 mM sodium phosphate (pH 7.2) containing 1 M NaCl. This eluate was collected and concentrated using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cutoff of 8,000; manufactured by Filtron), thereby obtaining a flow-through fraction Fl (460 ml; protein concentration, 2.11 mg/ml; total protein 973 mg; relative activity 16.3). Then the elution solution was changed to 2M NaSCN and the thus eluted TSK-gel AF-BLUE 650MH-adsorbed TPO-active fraction F2 (2840 ml) was concentrated to 6.81 ml using an ultrafiltration unit with YM 3 membrane (76 mm diameter, manufactured by Amicon Corp.). Total protein in this TPO-active fraction F2 was 12.5 mg and protein yield of F2 in this step was 0.62%. Relative activity of TPO was calculated to be 240. Then F2 was loaded onto a HiLoad 26/60 Superdex 200 pg (manufactured by Pharmacia Biotech, catalog no. 17-1071-01; 2.6 cm in diameter and 60 cm material height) and developed with 20 mM sodium acetate (pH 5.5) containing 50 mM NaCl with a flow rate of 1 ml/min. After completion of the development, the TPO activity was found in the eluates in the range from 194 to 260 ml after application. These eluates were collected to obtain a Superdex 200 pg TPO-active fraction F2 (66 ml; protein concentration, 0.012 mg/ml; total protein 7.41 mg; relative activity, 142,860; total activity, 1,058,600). Then buffer A (20 mM sodium acetate, pH 5.5) and buffer B (20 mM sodium phosphate, pH 7.2, containing 500 mM NaCl) were prepared and the TPO-active fraction was set at a flow rate of 1 ml/min. on a strong cation exchange column RESOURCE S (manufactured by Pharmacia Biotech, catalog no. 17-1178-01; 0.64 cm diameter and 3 cm material height) previously equilibrated with 100% A. Then the elution was carried out at a flow rate of 0.3 mg/min . with 40 minutes linear gradient from 100% A to 100% B. When alaysed, the TPO activity was detected in a wide eluate range (from 5% B to 32% B). These TPO-active eluates were collected and concentrated using an ultrafiltration unit with a YM 3 membrane (25 mm diameter, manufactured by Amicon Corp.) to obtain a RESOURCE S TPO-active fraction F2 (1.65 ml; protein concentration 4.74 mg/ml; total protein 7.82 mg; relative activity 71,400; total activity 558,600).

The partially purified TPO sample was administered subcutaneously for 5 consecutive days (474 mg (100 µl)/mouse/day) to male ICR mice (9 weeks old) whose platelet counts had been measured the day before administration. As a control, BSA (200 µl (100 µl)/mouse/day) or a buffer used as a solvent of the partially purified TPO (100 µl/mouse/day) was administered in the same manner. On the day after the last administration, blood was collected from the heart of each mouse to measure platelet count using a hemocytometer (F800, manufactured by Toa Iyo Denshi). In the mice administered partially purified TPO, an increase in platelet count by a factor of about 2.14 was observed compared to the platelet count before administration. When compared to the control groups, increased platelet counts in the partially purified TPO-administered mice were about 1.74 times more than the BSA-administered mice and about 1.90 times higher than the buffer-administered mice. These results showed that TPO promotes platelet production in vivo. In addition, where the amount of immunosuppressive acidic protein (IAP), known as a murine acute phase protein, was not only delivered in the mice adrened with partially purified TPO, this strongly indicated that TPO is different from IL-6 and IL-11 in the induction of acute phase protein.

(Example 12)

Detection of TPO mRNA in rat tissue

To localize rat TPO mRNA expressing tissues in the rat body, RNA was extracted

from different rat tissues. A total of 6 rats were exposed to X-ray radiation in the same manner as described in Example 1 and various tissues (brain, thymus, lung, liver, heart, spleen, small intestine, kidney, testis and bone marrow cells) were taken out from the rats on the 11th to 14th .day after X-ray irradiation and immediately frozen in liquid nitrogen. Extraction of total RNA was performed using an RNA isolation reagent ISOGEN (manufactured by Wako Pure Chemical Industries, Ltd.). The amount of ISOGEN to be used was determined according to the weight of each frozen tissue sample and the reagent-spiked tissue sample was treated with a homogenizer (PhyscotronR NS-60, manufactured by NITI-ON Medical & Physical Instruments MFG. Co., LTD) until complete disintegration of the tissue was obtained (about 45 to 60 seconds at 10,000 revolutions per minute). The tissue homogenate was then subjected to total RNA extraction using a procedure based on the acid guanidium phenol chloroform method according to Chomczynski et al. (Anal. Biochem., vol. 162, pp. 156-159, 1987). As a result, 1.1 to 5.6 mg of total mRNA was obtained from the respective tissues.

Using OligotextM-dT20 (Super) (manufactured by Japan Synthetic Rubber/Nippon Roche), 20 ng of poly (A) + RNA was purified from about 500 ng of total RNA.

Using random primers, the first strand of cDNA was synthesized from 1 ng of poly (A)+ RNA prepared from each of the tissues. That is, 1 ng of poly (A)+ RNA was dissolved in 10 ni sterile water, incubated at 70°C for 15 minutes then quickly cooled down. To this was added 75 pmol random primer (takara Shuzo Co., Ltd.), 10 U RNase inhibitor (Boehringer-Mannheim Corp.), 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2 and 200 U Super ScriptTM II (a reverse transcriptase manufactured by Life Technologies). The thus prepared solution (total volume 20 ni) was incubated at 37°C for 1 hour and the resulting reaction solution was incubated at 70°C for 10 minutes to inactivate the enzyme and stored at -20°C until its use.

New primers for PCR were synthesized based on the rat TPO cDNA sequence obtained in example 10. Sequences of the thus synthesized primers are as follows.

rTPO-L 5'-CCT GTC CTG CTG CCT GCT GTG-3' (SEQ ID NO: 33)

(positions 347 to 367 of SEQ ID NO: 2)

rTPO-N: 5'-TGA AGT TCG TCT CCA ACA ATC-3' (SEQ JD NO: 34)

(anti-sense primer corresponding to position 1005 to 1025 in SEQ ID NO: 2).

By using 1/10 volume of each of the synthesized cDNA solutions as template, 1 uM of each of thus synthesized rTPO-I and rTPO-N as primers and GeneAmp™ PCR reagent kit with AmpliTaq™ DNA polymerase (manufactured by Takara Shuzo Co ., Ltd.), PCR was performed in a volume of 100 µl using a GeneAmp™ PCR system 9600 (manufactured by PERKIN-ELMER) and heating at 95°C for 2 minutes, repeating a total of 30 cycles, each cycle consisting of denaturation at 95°C for 1 minute, annealing at 57°C for 1 minute and synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes. When each of the reaction solutions thus prepared was subjected to electrophoresis using 2% agarose gel (manufactured by FMC BioProducts) and amplified bands were examined, bands that appeared to be specific over TPO mRNA were found on gels resulting from brain, liver, small intestine and kidney. These results indicate that expression of rat TPO mRNA occurs in these organs, although its expression level cannot be assessed. Any possible existence of a corresponding expression mode in the human body and the results in example 4 are considered together, it seems that the liver is a suitable starting material for the acquisition of human TPO cDNA.

(Example 13)

Construction of human normal liver-derived cDNA library

Based on the results from Example 12, the liver was chosen as the starting material for the application of human TPO cDNA cloning. In the same manner as described in Example 7, double-stranded cDNA with an EcoRI recognition site at its 5' end and a Noti recognition site at its 3' end was synthesized from 5 µg of commercial normal human liver-derived poly(A) + RNA (prepared by Clontech; a product extracted based on the acid guanidinium rphenol chloroform method of Chomczynski et al.,) using the Time SaverTM cDNA Synthesis Kit (manufactured by Pharmacia) and the DIRECTIONAL CLONING TOOLBOX (manufactured by Pharmacia). The thus synthesized cDNA was ligated with 1.2 µg of the above-mentioned expression vector pEF18S which had been cut in advance with EcoRI and Noti and transformed into 8.4 ml of the above-mentioned Competent High E. coli DH5 (manufactured by Toyobo Co. , Ltd.). As a result, 1.2 x 10 6 transformants were obtained.

(Example 14)

Production (cloning) of human TPO cDNA fragment by PCR

Using random primers, the first strand of cDNA was synthesized from 1 ng of commercial normal human liver-derived poly(A)+ RNA (manufactured by Clontech). That is, 1 µg of poly(A)+ RNA was dissolved in 10 µl of sterile water, incubated at 70°C for 15 minutes and then rapidly cooled. To this was added 75 pmol random primer (Takara Shuzo Co.„ Ltd.), 10 U Rnase inhibitor (Boehringer Mannheim Corp), 50 mM Tris-HCl (pH 8.3), 75 mM KC1, 3 mM MgCl2 and 200 U of reverse transcriptase Super ScriptTM II (manufactured by Life Technologies). The thus prepared solution (total volume 20 µl) was incubated at 37°C for 1 hour and the resulting reaction solution was incubated at 70°C for 10 minutes to inactivate the enzyme and stored at -20°C until its use.

Primers for PCR use were synthesized based on the rat TPO cDNA sequence (SEQ ID NO: 2). Sequences of the thus synthesized primer are as follows:

rTPO-AlN: 5'-ATG GAG CTG ACT GAT TTG CTC-3' (SEQ ID NO: 35)

(positions 173 to 193 of SEQ ID NO: 2)

rTPO-N: 5'-TGA AGT TCG TCT CCA ACA ATC-3' (SEQ ID NO: 36)

(anti-sense primer corresponding to positions 1005 to 1025 in SEQ ID NO: 2).

Using 1/10 volume of the synthesized cDNA solution as template, 1 µM of each of these synthesized rTPO-AIN and rTPO-N as primers and GeneAmp™ PCR reagent kit with AmpliTaq™ DNA polymerase (manufactured by Takara Shuzo Co.) , PCR was performed in a volume of 100 µl using the GeneAmp™ PCR system 9600 (manufactured by PERKIN-ELMER) and heating at 95°C for 2 min, repeated for a total of 35 cycles, each cycle including denaturation at 95°C in 1 minute, annealing at 40°C for 1 minute and synthesis at 72°C for 1 minute followed by incubation at 72°C for 7 minutes.

The reaction solution thus obtained was subjected to 2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a DNA fragment of about 620 bp and the main product from PCR which was then purified using the above Prep-A-Gene DNA purification kit (manufactured by Bio-Rad Laboratories, Inc.). The nucleotide sequence and the thus purified DNA fragment were directly determined on a 373A DNA sequencer (manufactured by Applied Biosystems) using the above-mentioned Taq Dye Deoxy™ Terminator Cycle Sequencing Kit (manufactured by Applied Biosystems). The nucleotide sequence thus determined, excluding the primer part and an amino acid sequence derived therefrom, is shown in the sequence listing (SEQ ID NO: 3).

This DNA fragment, excluding the primer sequence, has a length of 580 bp. By comparison, the human cDNA shows a homology of 86% with the rat cDNA nucleotide sequence indicating that this DNA fragment encodes part of the human TPO cDNA.

(Example 15)

Screening of human TPO cDNA clone by PCR

Human TPO-corresponding primers from PCR were synthesized based on SEQ DD NO: 3. Sequences of the thus synthesized primers follow.

hTPO-I: 5'-TTG TGA CCT CCG AGT CCT CAG-3' (SEQ DD NO: 37)

(positions 60 to 80 of SEQ DD NO: 3)

hTPO-J: 5'-TGA CGC AGA GGG TGG ACC CTC-3' (SEQ DD NO: 38)

(anti-sense primer corresponding to positions 479 to 499 in SEQ DD NO: 3).

The human cDNA library constructed in Example 13 was amplified and divided into stocks (each stock containing about 100,000 clones), grown overnight in 1 ml of the above LB medium containing 50 mg/ml ampicillin also subjected to plasmid DNA extraction using a automated plasmid isolation apparatus PI-100 (VER-3.0, manufactured by Kurabo Industries, Ltd.). The thus extracted DNA was dissolved in TE solution.

Using 5% of the extracted DNA sample as template, 1 µm each of the synthesized oligonucleotides (hTPO-I and hTPO-J) as primer and GeneAmp TM PCR Reagent Kit with AmpliTaq™ DNA polymerase (manufactured by Takara Shuzo Co. , Ltd.), PCR was performed in 20 µl volume by GeneAmp™ PCR system 9600 (manufactured by PERION-ELMER) (35 cycles in total, each cycle includes denaturation at 95°C for 1 minute, annealing at 59°C for 1 minute and synthesized at 72°C for 1 minute followed by final incubation at 72°C for 7 minutes). As a result, a specific band was detected in 3 of the 90 stocks used. One of these pools was divided into subpools, each containing about 5000 clones and plasmid DNA was purified from 90 subpools before PCR in the same way. As a result, specific bands were detected in 5 sub-repositories. One of these 5 stocks was divided into sub-stocks each containing 250 clones and plasmid DNA was extracted from 90 sub-stocks. When these extracted samples were subjected to PCR in the same way, a specific band was detected in 3 sub-pools. One of these three stocks was further divided into sub-stocks, each containing 30 clones and plasmid DNA was purified from 90 sub-stocks to perform PCR in the same way. As a result, a specific band was detected in 3 sub-repositories. One of these current supplies was cultured to the above-mentioned LB plate containing 50 µg/ml ampicillin and each of the colonies thus formed was subjected to plasmid DNA extraction and PCR in the same way. As a result, a clone HL34 is finally obtained.

(Example 16)

Sequencing of TPO cDNA

Purification of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). The clone HL34 was grown overnight in 50 ml of LB medium containing 50 µg/ml ampicillin and the resulting cells collected by centrifugation were suspended in 4 ml of the above TEG-lysozyme solution. To this was added 8 ml of 0.2 N NaOH/1% SDS solution and 6 ml of 3 M potassium/5 M acetate solution to thoroughly suspend the cells. After centrifugation of the suspension, the resulting supernatant was treated with phenol/chloroform (1:1) mixed with the same volume of isopropanol and then centrifuged. The resulting pellet was dissolved in TE solution and treated with RNase and then with phenol/chloroform (1:1), followed by ethanol precipitation. The resulting pellet was again dissolved in TE solution to which NaCl and polyethylene glycol 3000 are then added to a final concentration of 0.63 M and 7.5% respectively. After centrifugation, the pellet was dissolved in TE solution and washed with ethanol. In this way, about 300 µg of pEF18S-HL34 plasmid DNA was obtained.

The thus purified DNA was loaded onto the above 373A DNA sequencer (manufactured by Applied Biosystems) to determine its complete nucleotide sequence using the above Taq Dye Deoxy™ Terminator Cycle Sequencing Kit (manufactured by Applied Biosystems). The nucleotide sequence thus determined and the amino acid sequence derived therefrom are shown in the sequence listing (SEQ ID NO: 4). In this

case, oligonucleotides were synthesized based on the nucleotide sequence of SEQ JD

in NO: 3 and synthetic oligonucleotides designed based on the internal sequence obtained by sequence analysis, used in the nucleotide sequence determination as primers.

As a result, it was confirmed that the plasmid clone pEF18S-HL34 comprised a cDNA fragment of 861 bp and which contained highly homologous sequences with the amino acid sequence AP8 (amino acid No. 1 to 12 of SEQ ID NO: 4) and TP2/TP3 (amino acid Nos. 157 to 162 of SEQ ID NO: 4) analyzed in Example 2. This DNA fragment appears to encode an open reading frame at its position 25, but has no termination codon and contains a poly A tail-like sequence of 76 bases on its 3' end. Its

amino acid sequence, consisting of 253 amino acid esters just before a poly A tail-like in sequence, showed 84% homology with the corresponding part of rat TPO cDNA (an amino acid sequence part consisting of 147 residues shown in SEQ ID NO: 2). Accordingly, this clone was found to be a DNA fragment encoding a portion of human cDNA corresponding to rat TPO cDNA. Due to the absence of a termination codon and the presence of a poly A tail-like sequence at its 3' end, this clone appeared not to be a complete DNA, but an artificial product arising during the cDNA library construction procedure.

(Example 17)

Expression of human TPo cDNA in COS 1 cells and confirmation of the TPO activity Transfection of the thus obtained plasmid clone pEF18S-HL34 into COSI cells was carried out according to the procedures in example 11. That is, transfection was carried out with 10 µg of plasmid DNA by the DEAE-dextran method which includes chloroquine treatment. The COS 1 cells were cultured for 3-5 days at 37°C and then the supernatant was collected.

The culture supernatant thus obtained was thoroughly dialyzed against IMDM culture medium and evaluated with the rat CFU-MK analysis system. TPO activity was detected in a dose-dependent manner in the culture supernatant in COSI cells where a plasmid pEF18S-HL34 was expressed (fig. 8). After 4 days of culture, many megakaryocytes formed elongated cytoplasmic processes. In contrast, TPO activity was not found in the culture supernatant of COSI cells in which plasmid pEF18S alone was expressed (fig. 8). In the M-07e assay system, the M-07e cell proliferative seeking activity was found only in the culture supernatant of COS 1 cells in which plasmid pEF18S-HL34 was expressed through its transfection. These results showed that pEF18S-HL34 contains a gene that codes for a protein with TPO activity. In addition, human TPO was shown to act on rat megakaryocytic progenitor cells (no species specificity).

(Example 18)

Expression of human TPO deletion tvpe cDNA and confirmation of TPO activity The clone HL34 obtained in Example 15 contained cDNA comprising a poly A-like continuous sequence at its 3'-tagged side, which did not exist in rat TPO cDNA and therefore appeared to be an artificial product of the experiment. Accordingly, a cDNA molecule from which the poly A tail-like sequence has been deleted was prepared to see that a protein expressed by the deletion cDNA had a TPO activity. Deletion cDNA was prepared using PCR. Sequences of primers used for PCR are as follows, where a restriction enzyme recognition site is added at the 5' end of each primer (EcoRI to

hTP05 and Noti and two stop codons TAA and TGA for hTP03).

hTP05: 5' TTT GAA TTC GGC CAG CCA GAC ACC CCG GCC-3' (SEQ ID NO:

170) (positions 1 to 21 of SEQ ID NO: 4)

hTP03: 5'-TTT GCG GCC GCT CAT TAT TCG TGT ATC CTG TTC AGG TAT CC-3' (SEQ ID NO: 171) (anti-sense primer corresponding to position 757 to 780 in SEQ ID NO: 4).

Using 1 µg of plasmid DNA from the plasmid clone pEF18S-HL34 obtained in Example 16 as template, 10 µm of each of these synthesized hTP05 and hTP03 as primers and the GeneAmp<TMp>CR Reagent Kit with AmpliTaq<TM> DNA polymerase (prepared by Takara Shuzo Co., Ltd.), PCR was performed in a volume of 100 µl using the GeneAmp™ PCR System 9600 (manufactured by PERKLN-ELMER) and repeating a total of 15 cycles, each cycle comprising denaturation at 95°C in 1 minute, annealing at 65°C for 1 minute and synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes. A band of about 800 bp thus prepared was cut with restriction enzymes EcoRI and Noti, purified and subcloned into the expression vector pEF18S which had been previously treated in the same way with restriction enzymes. From the resulting transformants, clones containing a DNA fragment of about 800 bp were selected to prepare large amounts of plasmid DNA according to the procedure described in Example 5. The entire length of the amplified region of about 800 bp of each plasmid was subjected to nucleotide sequence analysis to find its complete identity with the nucleotide sequence analyzed in Example 16 (positions 1 to 780 of SEQ ID NO: 4).

The plasmid clone thus produced was named pHT 1-231. The vector pHT 1-231 originates from an E. coli strain DH5 has been deposited by the present inventors on February 14, 1994 under deposit number FERM BP-4564, in the National Institute of Bioscence and Human Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

Transfection of the thus obtained plasmid into COSI cells was carried out according to the procedure in Example 11. That is, transfection was carried out with 10 µg of plasmid DNA by the DEAE-dextran method which includes chloroquine treatment. The COSI cells were cultured for 3 to 5 days at 37° C and the supernatant collected.

The thus prepared culture supernatant was thoroughly dialyzed against IMDM culture medium and evaluated with the rat CFU-MK analysis system. The TPO activity was found in the culture supernatant in COSI cells in which plasmid pHT 1-231 was expressed (fig. 9). On the fourth day of culture, many megakaryocytes formed elongated cytoplasmic processes. In contrast, TPO activity was not found in the culture supernatant of COSI cells where the plasmid pEF18S alone was expressed (fig. 9). In the M-07e analysis system, M-07e cell proliferation-seeking activity was found only in the culture supernatant of COSI cells where the plasmid pHT 1-231 was expressed. These results show that pHT 1-231 contains a cDNA that codes for a protein that has TPO activity.

(Example 19)

Preparation of the 3'-tagged end region of human TPO cDNA by PCR

The clone HL34 prepared in Example 15 contained cDNA comprising the poly(A) tail-like sequence such that the predicted 3' end region was incomplete. Attempts are therefore made to obtain a full-length 3'-end region by PCR. The following four types of 5' primers for PCR were synthesized on the basis of the sequences found in Example 16

hTPO-H: 5'-AGC AGA ACC TCT CTA GTC CTC-3' (SEQ ID NO: 39)

(positions 574 to 594 of SEQ ID NO: 4)

hTPO-K: 5'-ACA CTG AAC GAG CTC CCA AAC-3' (SEQ ID NO: 40)

(positions 595 to 615 of SEQ ID NO: 4)

hTPO-N: 5'-AAC TAC TGG CTC TGG GCT TCT-3' (SEQ ID NO: 41)

(positions 660 to 680 of SEQ ID NO: 4)

hTPO-O: 5'-AGG GAT TCA GAG CCA AGA TTC-3' (SEQ ID NO: 42)

(positions 692 to 712 of SEQ JD NO: 4).

Following 3-sided primers containing mixed nucleotides at the four bases of the 3'-tagged end were synthesized to amplify cDNA from the start of the poly(A) portion. An anchor primer without the scrambled nucleotide was also synthesized.

The first strand cDNA was synthesized from 1 µg of commercial normal liver-derived poly(A)<+>RNA (manufactured by Clontech), as in Example 14, but using 0.5 µg oligo dT primer (it is included in the TimeSaver™ cDNA Synthesis Kit manufactured by Pharmacia). Using 1/10 volume of late synthesized cDNA solution as a template, 20 µM hTPO-H primer and 10 µM hTP03 mix primer and GeneAmp PCR reagent kit with AmpliTaq TM DNA polymerase (manufactured by Takara Shuzo Co., Ltd), PCR was performed in a volume of 50 µl using the GeneAmp™ PCR system 9600 (manufactured by PERKIN-ELMER) and heated to 96°C for 2 minutes, repeating a total of 10 cycles, each cycle comprising a heat denaturation at 96°C for 1 minute, annealing at 48°C for 1 minute and synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.

The second PCR was performed using 1/10 volume of the resulting solution from the first PCR as a template, 20 [mu]m hTPO-K primer and 10 [mu]m of the hTP03mix primer with volume pp 50 [mu]l, heating to 96°C for 2 minutes, repeating a total of 10 cycles, each cycle consisting of heating denaturation at 96°C for 1 minute, annealing at 63°C for 1 minute and synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.

The third PCR was performed using 1/10 volume of the resulting solution from the second PCR as a template, 20 nM hTPO-N primer and 10 µm hTPO3 mix primer with a volume of 50 µl and heating at 96°C for 2 minutes, repeating a total of 10 cycles, each cycle comprising heat denaturation at 96°C for 1 minute, annealing at 63°C for 1 minute and a synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.

The fourth PCR was performed using 1/10 volume of the resulting solution from a third PCR as template, 20 hTPO-0 primer and 10 nM hTP03 mix primer in volume of 50 [mu]l and heating at 96°C for 2 minutes, repeat of a total of 10 cycles, each cycle comprising heat denaturation at 96°C for 1 minute, annealing at 63°C for 1 minute and a synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.

The fifth PCR was performed using 1/10 volume of the resulting solution from a fourth PCR as a template, 20 nM hTPO-0 primer and 20 nM hTPO3 anchor primer with a volume of 50 ni and heating at 96°C for 2 minutes, repeating a total of 10 cycles, each cycle comprising heat denaturation at 96°C for 1 minute, annealing at 63°C for 1 minute and a synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.

The reaction solution thus obtained was subjected to 2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a DNA fragment of about 600 bp as the main product of PCR which was then purified using the above-mentioned Prep-A gene DNA resection kit (manufactured by BioRad Laboratories, Inc.). Nucleotide sequence of the thus purified DNA fragment was then determined by means of a 373A DNA sequencer (manufactured by Applied Biosystems) using the above-mentioned Taq Dye Deoxy Terminaler Cycle Sequencing Kit (manufactured by Applied Biosystems). The nucleotide sequence thus determined and the amino acid sequence derived therefrom are shown in the sequence list (SEQ ID NO: 5).

This DNA fragment has a nucleotide sequence coding for 130 amino acids starting with the primer hTPO-0 followed by a sequence of more than 180 nucleotides with the 3' end (the nucleotides after position 577 could not be determined). The 30 amino acid sequence starts with glycine which matches the amino acid sequence at positions 203 to 232 of SEQ ID NO: 4. The nucleotide sequence at positions 1 to 94 also matches the nucleotide sequence at positions 692 to 785 of SEQ ID NO: 4.

As a result of the sequence analysis of the cDNA fragment in Example 17 and the PCR amplified fragment in the present example, it is reasonable to estimate that human TPO protein consists of 353 amino acids containing a 21 amino acid signal sequence as shown in

SEQ DD NO: 6.

(Example 20)

Reconstruction of human normal liver-derived cDNA library

As the clone HL34 obtained in Example 15 contained the poly A tail-like sequence directly at its 3' end of its open reading frame without a termination codon and therefore appeared to be a favorable product of the cDNA synthesis, a cDNA library was reconstructed using 5 ng of commercial normal human liver-derived poly

(A)<+> RNA preparation (manufactured by Clontech). Synthesis of cDNA was performed using the SuperScript™ Lambda System for cDNA synthesis and ' Cloning Kit and SuperScript<TM> jj

RNase H" (both manufactured by LIFE TECHNOLOGIES). Poly (A)<+> RNA was subjected to heat denaturation and then added 20 µl of a reaction solution containing a Noti sequence-included oligo dT as a primer attached to the kit (50 mM Tris- HCl, pH 8.3,75 mM KCl, 3 mM MgCl2,1 mM DTT, 1 mM dNTP mix, 200 U SuperScript™ H RNase H~) followed by 60 min incubation at 37° C. After synthesis of the second cDNA strand (2 hours incubation at 16°C in 150 ul of a reaction solution containing 25 mM Tris-HCl, pH 8.3, 100 mM KCl, 5 mM MgCl2, 250 nM dNTP mix, 5 mM DTT, 40 U of E. coli DNA polymerase I, 2 TJ of E. coli RNase H and 10 U of E. coli DNA ligase), 10 TJ of T4 DNA polymerase were added and the resulting mixture was incubated at 16°C for 5 minutes. The reaction solution was heated at 65°C for 10 hours and extracted with the same volume of phenol/chloroform, and cDNA molecules with a length of less than 400 bp were removed using SizeSep^2M 400 spun column (a spun column for removal of low molecular weight DNA, connected to the TimeSaver TM cDNA synthesis kit manufactured by Pharmacia). After addition of EcoRI adapter (linked to Directional Cloning Toolbox manufactured by Pharmacia), the thus treated sample was cut with Noti and again taken on SizSep™ 400 spun column to remove low molecular weight DNA. The thus synthesized 1.3 ng double-stranded cDNA having an EcoRI recognition sequence at its 5'-end and Noti recognition sequence at its 3'-end was ligated with the expression vector pEF18S which has been previously cut with EcoRI and Noti and then transformed into 9 .2 ml Competent High E. coli DH5 (manufactured by Toyobo Co., Ltd.). The thus produced human cDNA library (hTPO-Fl) contained 1.0 x 10<6> transformants.

(Example 21)

Screening of TPO cDNA cloned from human liver cDNA library hTPO-Fl

Human TPO cDNA corresponding primers from PCR were synthesized based on sequence ID No. 3 and 6 shown in the sequence listing. The sequences of the thus synthesized primers are as follows.

hTPO-I: 5'-TTG TGA CCT CCG AGT CCT CAG-3' (SEQ ID NO: 48)

(positions 60 to 80 of SEQ ID NO: 3)

hTPO-KU: 5'-AGG ATG GGT TGG GGA AGG AGA-3' (SEQ ID NO: 49)

(anti-sense primer corresponding to the sequence in position 901 to 921 in SEQ ID NO: 6).

Human liver cDNA library hTPO-F1 (1.0 x 10*> transformants) constructed in Example 20 was divided into three pools (pools No. 1 to 3) and the pools were frozen. PCR was performed using 2 µm of plasmid DNA prepared from each of the stocks as a template and 1 µm of each of the synthesized oligonucleotides (hTPO-I and hTPO-KU) as primers. Using GeneAmp™ PCR Reagent Kit with AmpliTaq™ DNA polymerase (manufactured by Takara Shuzo Co., Ltd.), PCR was performed in 100 µl volume in GeneAmp™ PCR system 9600 (manufactured by PERKIN-ELMER) (a total of 35 cycles , each cycle consisting of denaturation at 95°C for 1 min, annealing at 59°C for 1 min and synthesis at 72°C for 1 min followed by final incubation at 72°C for 1 min). As a result, a DNA fragment of the expected size was amplified when the plasmid DNA prepared from stock No. 3 was used. Stock No. 3 was divided into substocks each containing 15,000 transformants and these substocks were grown overnight in 1 ml LB medium containing 50 µg/ml ampicillin followed by extraction of plasmid DNA using automated plasmid isolation apparatus PI-100. The thus extracted DNA was dissolved in TE solution and 5% of the resulting solution was used as a template when carrying out PCR using the same primers and under the same conditions as described above. As a result, amplification of cDNA of the expected size was found in 6 of the 90 pools. When one of these positive pools was divided into subpools, each containing 1000 clones and plasmid DNA was extracted and PCR was performed in the same manner as described above, DNA amplification was not observed. The density of the bands on electrophoresis gels of DNA fragments amplified by a series of PCRs became thinner as the subpools were further divided, which appeared to be caused by low recovery of plasmid DNA due to poor growth of the clone in question. From there, a second screening was performed by colony hybridization using the original No. 3 stock.

No. 3 was spread on 100 LB agar plates of 15 cm diameter with an inoculum size such that 4100 colonies grew on each LB agar plate. After preparing a replicate plate from each of the plates thus inoculated, one of the duplicate plates was cultured at 37°C for 6 hours to recover colony growth on the plate and to extract the plasmid DNA samples. When PCR in these DNA samples was performed in the same manner as described above, amplification of a band equivalent to the expected size was observed in one of 100 subpools. Two replica filters were prepared from the slab of this substock using BIODYNE™ EN A TRANSFER MEMBRANE (manufactured by PALL). Denaturation of the filters was performed by developing them in 10% SDS for 10 min, in 0.5 N NaOH/1.5 M NaCl for 10 min and then 0.5 M tris-HCl (pH 8.0)/1.5 M NaCl for 10 min in the order, followed by 30 minutes in a vacuum oven. The thus coated filters were washed with 6 x SSC (prepared by diluting 20 x SSC stock solution comprising 175.3

grams of NaCl and 88.2 g of Na citrate dissolved in 1 liter of water, pH 7.0) with 1% SDS added.

i Prehybridization of the thus washed filters was carried out by incubating it at 42°C for 30 minutes with shaking in 30 ml and reaction solution comprising 50% formamide, 5 x SSC, 5 x Denhardt's solution (prepared from 50 x Denhardt's solution containing 5 g Ficoll, 5 g polyvinylpyrrolidone and 5 g bovine serum albumin fraction V in 500 ml water), 1% SDS and 20 µg/ml salmon sperm DNA. After prehybridization, the reaction solution was exchanged with 30 ml of a hybridization solution of the same composition, mixed with a probe that had been labeled with [6- <32p>]dCTP (manufactured by Amersham) and then incubated with shaking at 42°C for 20 h . The labeled probe used in this experiment was EcoRI/BamHI fragment of plasmid

pEF18S-HL34, which had been obtained by purifying part of SEQ ID NO: 4

in exciting from its 5-end to the base at the 458 position and by labeling the purified portion by random primer technique with the megaprime DNA Labeling System (a kit manufactured by Amersham based on the method described in Anal. Biochem., 132, 6-13, 1983). The thus treated filters were washed in 2 x SSC/0.1% SDS solution at 42°C for 30 minutes and then in 0.2 x SSC/0.01% SDS solution at 42°C for 30 minutes. The filters were then subjected to 16 hours of autoradiography at -70°C using an intensifying screen and X-OMAT® AR5 film (manufactured by Eastman Kodak). As a result, a single signal considered to be positive was observed. Colonies approximately corresponding to this signal were picked from the original plate

inoculated again into a 10-cm LB agar plate. A total of 50 colonies were grown on the surface

in separately cultured and their DNA samples were subjected to PCR using the above-mentioned primers hTPO-I and hTPO-KU under the same conditions as described above. As a result, amplification of a band equivalent to the expected size was found in only one clone, which was designated pHTF1.

(Example 22)

Determination of nucleotide sequence of human TPO cDNA cloned pHTF1

Purification of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). The clone pHTF1 was grown overnight in 50 ml of the LB medium containing 50 µg/ml ampicillin. The resulting cells were collected by centrifugation and suspended in 4 ml of the above TEG-lysozyme solution. To this was added 8 ml of 0.2 N NaOH/1% SDS solution and then 6 ml of 3 M potassium/5 M acetate solution to thoroughly suspend the cells. After centrifugation of the suspension, the resulting supernatant was treated with phenol/chloroform (1:1), mixed with the same volume of isopropanol and then centrifuged. The resulting pellet was dissolved in TE solution treated with RNase and then with phenol/chloroform (1:1), followed by ethanol precipitation. The resulting pellet was again dissolved in TE solution to which NaCl and polyethylene glycol 3000 were then added to their final concentration of 0.63 M and 7.5% respectively. After centrifugation, the pellet was dissolved in TE solution and washed with ethanol. In this way, 300 µg of plasmid DNA pHTF1 was obtained.

The thus purified plasmid DNA was loaded onto the above 373A DNA sequencer (manufactured by Applied Biosystems) to determine its complete nucleotide sequence using the above Taq Dye Deoxy<TM> Terminaler Cycle Sequencing Kit (manufactured by Applied Biosystems). The nucleotide sequence thus determined and the deduced amino acid sequence were shown in sequence listing (SEQ ID NO: 7). The oligonucleotide synthesis based on nucleotide sequence SEQ DD NO: 6 and the internal sequence obtained by their sequence reaction analysis were used in the nucleotide sequence determination as primers.

As a result, it was confirmed that the clone pHTF1 contains a cDNA fragment of 1,721 bp and a high homology with amino acid sequence AP8 (amino acids 1-12 of SEQ DD NO: 7) and TP2/TP3 (amino acid sequence Nos. 157 to 162 of SEQ DD NO : 7) analyzed in Example 2. This DNA fragment appeared to have a 5' non-coding region pp 101 bases, an open reading frame comprising 353 amino acid residues, starting with a methionine residue grafted at 102 to 104 nucleotide positions and terminated by a glycine residue encoded at 1158 to the 1160 nucleotide positions, the following termination codon (TAA), a 3' non-coding region of 531 bases and a poly A neck sequence of 30 bases. The amino acid sequence of a protein thought to be encoded by the open reading frame completely matched the predicted amino acid sequence of human TPO shown in SEQ DD NO: 6. The cDNA sequence of pHTF1 has a larger size than the cDNA sequence shown in SEQ DD NO: 6 and contains 77 additional bases at the 5' side and 347 additional bases at the 3' side in front of its poly

A tail sequence. The nucleoside sequence also differs from SEQ DD NO: 6 at 3

positions. That is, A (position 84), A (position 740) and G (position 1,198) in SEQ DD NO: 7 where respectively C, T and A in SEQ DD NO: 6. Only the mutation in position 740 was included in the protein -coding region, but this mutation did not cause any amino acid substitution due to both bases A and T being the third base of

threonine codone. Nevertheless, the reason for these base substitutions was not clear at the time,

the analysis of the plasmid clone pHTF1 confirmed that human TPO protein comprises 353 amino acid residues with 21 amino acid signal sequences. The molecular weight of the mature protein excluding the signal sequence was estimated to be 35,466.

Vector pHTF1 carried by an E. coli strain DH5 has been deposited by the present inventors on March 24, 1994 under accession no. FERM BP-4617, at the National Institute of Bioscience and Human-Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

(Example 23)

The expression of human TPO cDNA cloned pHTFl in COS 1 cells and confirmation of TPO activity

Transfection into COS 1 cells had of the thus obtained plasmid clone pHTF1 was carried out according to the procedure in example 11. That is, transfection was carried out with 10 µg of plasmid DNA by the DEAE-dextran method which includes chloroquine treatment. The transfected COS 1 cells were cultured for 3 days at 37°C and the culture supernatant was collected.

The culture supernatant was thoroughly dialyzed against IMDM culture medium and evaluated by the rat CFU-MK analysis system. TPO activity was detected in a dose-dependent manner in

the supernatant in COS 1 cells in which plasmid pHTF1 was expressed (fig. 10a). In contrast, TPO activity was not found in the culture supernatant of COS 1 cells where the plasmid pEF18S alone was expressed (fig. 10a). Corresponding results were obtained in the M-07e analysis system. The supernatant from the COS 1 cells in which the plasmid pHTF1 was expressed was expressed

significant M-07e cell proliferation in a dose-dependent manner (Fig. 10b). These results showed that pHTFl contains a gene that codes for a protein with TPO activity.

(Example 24)

Cloning of human TPO chromosomal DNA

Cloning of human TPO chromosomal DNA was performed using human TPO cDNA as a probe. A genomic library used in the cloning was a gift from Professor T. Yamamoto at the Gene Research Center, Tohoku University (the library reported by Sakai et al. in Biol. Chem., 269,2173-2182,1994, constructed by partial cutting of human chromosomal DNA with a restriction enzyme Sau3Al and ligation of the partially cut to the BamHI site on the phage vector lambda EMBL3 prepared by Stratagene). Screening from this library using human TPO cDNA as a probe was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). Using E. coli LE392 as host, the library was inoculated onto a 15 cm plate containing NZYM (10 g of NZ amine, 5 g NaCl, 5 g Bacto Yeast Extract, 2 g MgSC«4.7H2O and 15 g agar in 1 liter of water, pH 7.0) in such an inoculum size that one plate contained 30,000 phage particles. Two replica filters were prepared from each of the 18 plates thus prepared using BIODYNE™ A TRANSFER MEMBRANE (manufactured by PALL). Denaturation of the filters was performed by immersing them in 0.5 N NaOH/1.5 M NaCl for 10 min and then 0.5 M tris-HCl (pH 8.0)/1.5 M NaCl for 10 min followed by 30 min air drying and subsequent 1.5 h baking at 80°C in a vacuum oven. Prehybridization was carried out by incubating the thus treated filters at 42°C for 1 hour in 500 ml reaction solution containing 50% formamide, 5 x SSC, 5 x Denhardt's solution, 1% SDS and 20 [Ag/ml salmon sperm DNA. For use as a probe, a human TPO cDNA fragment (base position no. 178 to 1,025 in SEQ ID NO; 7) was amplified by PCR and purified and the purified fragment was labeled with <32>p using the Random Primer Labeling Kit ( a DNA tag set prepared by Takara Shuzo based on the random primer method described in Anal. Biochem., 132,6-13, (1983). Sequences of primers used in this PCR are as follows.

hTPO-I: 5'-TTG TGA CCT CCG AGT CCT CAG-3' (SEQ ID NO: 50)

(positions 178 to 198 of SEQ DD NO; 7)

hTPO-N: 5'-AGG GAA GAG CGT ATA CTG TCC-3' (SEQ DD NO: 51) (anti-sense primer corresponding to the sequences at positions 1.005 to 1.025 in SEQ DD NO: 7).

Using the isotopically labeled probe, hybridization was performed at 42°C for 20 hours in 500 ml reaction solution with the same composition as prehybridization solution. The resulting filters were washed three times in 2 x SSC/0.1% SDS solution at room temperature for 5 min and once in 0.1 x SSC/0.1% SDS solution at 68°C for 1 h. The filters were subjected to 16 hours of autoradiography at -70°C using an intensification screen and X-OMAT^<M> AR5 film (manufactured by Eastman Kodak). As a result, 13 positive signals were obtained. Plaques approximately corresponding to each of the positive signals were collected from the original plates and re-inoculated onto 15 cm NZYM plates to form an inoculum size of 1000 plaques on each plate. Two replicate filters were prepared from each of the resulting plates to perform hybridization under the same conditions as above. As a result, positive signals were detected on all filters in the 13 groups. A single plaque was recovered from each of the resulting plates to prepare phage DNA by the plate lysate method described in Molecular Cloning. Phage DNA samples thus prepared from 13 clones were checked for the presence of the cDNA coding region by PCR using the primer with the following sequences.

hTPO-L: 5'-GGC CAG CCA GAC ACC CCG GCC-3' (SEQ DD NO: 52)

(positions 1 to 21 of SEQ DD NO: 6)

hTPO-F: 5'-ATG GGA GTC ACG AAG CAG TTT-3' (SEQ DD NO: 53)

(anti-sense primer corresponding to the sequence in positions 127 to 147 of SEQ DD NO: 6)

hTPO-P: 5'-TGC GTT TCCTGA TGC TTG TAG-3' (SEQ ID NO: 54)

(positions 503 to 523 of SEQ JD NO: 6)

hTPO-V: 5'-AAC CTT ACC CTT CCT GAG ACA-3' (SEQ ID NO: 55)

(anti-sense primer corresponding to the sequence in positions 1,070 to 1,090 in SEQ ID NO: 6).

When PCR was performed using these primer combinations, 5 out of 13 appeared

clones to contain the entire amino acid coding region from the cDNA. Chromosomal DNA measured in these 5 clones had a corresponding length of about 20 kb and showed almost the same pattern when a preliminary restriction enzyme analysis was performed. Therefore, one of these clones (clone HGT1) was selected and analyzed by Southern blotting. That is, 1 µg of DNA from clone UGT1 was cut completely by restriction enzyme EcoRI or Hind Ul and put on 0.8% agarose gel electrophoresis and the bands on the gel were transferred onto BIODYNE™ A TRANSFER MEMBRANE (manufactured by PALL). The resulting filter was air dried for 30 minutes subjected to 2 hours of baking at 80°C in a vacuum oven. Prehybridization was performed by incubating the thus treated filter at 42°C for 1 hour in 50 ml reaction solution consisting of 50% formamide, 5x SSC, 5x Denhardt's solution, 1% SDS and 20 µg/ml salmon sperm DNA. For use as a probe, human TPO cDNA fragment (base position No. 178 to 1,025 in SEQ JD NO: 7) was amplified by PCR and purified and the purified fragment was labeled at <32>p Using the Random Primer DNA Labeling Kit (manufactured by Takara Shuzo ).Using the isotopically labeled probe, hybridization was performed at 42°C for 20 hours in 50 ml of a reaction solution with the same composition of the prehybridization solution. The resulting filter was washed three times in 2 x SSC/0.1% SDS solution with room temperature for 5 minutes also again in 0.1 x SSC/0.1% SDS solution at 68°C for 1 hour.The filter was then subjected to 16 hours of autoradiography at -70°C using an intensifying screen and X-OMAT^<M> ^ R5 (manufactured by Eastman Kodak).As a r result, a single band of about 10 kb was observed on HindJU cutting. Therefore, 10 µg of DNA from clones 'HGT1 was cut with HindJU and subjected to 0.8% agarose gel electrophoresis and 10 kb bands were excised from the gel, purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and subcloned into a cloning vector pUC13 (manufactured by Pharmacia) that had been previously cut with HindJU. In this case, Competent-High E. coli DH5 produced by TOYOBO was used as the host strain.

Of the clones thus obtained, a clone containing the 10 kb HindJU fragment was selected and designated pHGT1.

A restriction map of phage clone 'HGT1' is shown in fig. 11.

(Example 25)

Determination of nucleotide sequence of human TPO chromosomal clone pHGT1 Cultivation of clone pHGT1 and purification of plasmid DNA was performed professionally according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). The clone pHGT1 was grown overnight in 50 µl LB medium containing 50 µg/ml ampicillin. The resulting cells were collected by centrifugation and suspended in 4 ml of the above TEG-lysozyme solution. To this was added 8 ml of 0.2 N NaOH/1% SDS solution and then 6 ml of 4 M potassium/5 N acetate solution which thoroughly suspends the cells. After centrifugation of the suspension, the resulting supernatant was treated with phenol/chloroform (1:1), mixed with the same volume of isopropanol and then centrifuged. The resulting pellet was dissolved in TE solution and treated with RNase and then with phenol/chloroform (1:1), followed by ethanol precipitation. The resulting pellet was again dissolved in TE solution to which NaCl and polyethylene glycol 3000 were added to the final concentration of 0.63 M and 7.5% respectively. After centrifugation, the pellet was dissolved in TE solution and washed with ethanol. In this way, about 300 µg of plasmid DNA pHGT1 was obtained.

Thus, using the above Taq Dye Deoxy<TM> Terminaler Cycle Sequencing Kit (manufactured by Applied Biosystems), it was purified by DNA put on the above 373A DNA sequencer (manufactured by Applied Biosystems) to determine its nucleotide sequence around the protein coding region from the cDNA nucleotide sequence. The nucleotide sequence thus determined and the deduced amino acid sequence are shown in the sequence listing (SEQ ID NO: 8). In this case, oligonucleotides used in Example 22 for nucleotide sequence analysis of cDNA and synthetic oligonucleotides formed based on the internal sequence obtained by their sequence reaction analysis were used as primers in the nucleotide sequence determination.

As a result, chromosomal DNA carried by the plasmid clone pGHT1 was found to contain the entire coding region of the amino acid sequence derived from SEQ ID NO: 6, and the nucleotide sequence of the coding region completely matched that of SEQ ID NO: 6.1 addition, a region corresponding to the amino acid coding exon 4 introns with lengths of 231 bp, 286 bp. 1,932 bp and 236 bp in that order from the 5-end (Fig. 11). The nucleotide sequence also differed from the cDNA nucleotide sequence in SEQ ID NO: 7 at 3 positions which were the same position described in example 22 (different positions between SEQ ID NO: 6 and 7). That is, A (position 84), A (position 740) and G (position 1,198) in SEQ ID NO: 7 were respectively C, T and A in SEQ ID NO: 8. Thus, it was revealed that the nucleotide sequence of the human TPO The cDNA clone pHTF1 obtained in Example 21 differs from the nucleotide sequence in the human chromosomal DNA clone pHGT1 in 3 positions. Therefore, to analyze whether these mutations in the cDNA clone pHGT1 sequence reflect the chromosomal DNA sequence, the nucleotide sequence of the other four clones among the five chromosomal DNA clones independently selected by screening, determined. The sequence analysis was carried out by the direct nucleotide sequence determination method using a phage DNA sample prepared according to the plate lysate method described in Molecular Cloning. Sequencing of each clone was carried out under the assumption that on the above-mentioned 373A DNA sequencer (manufactured by Applied Biosystems), using the sequencing primers used in Example 22 which had been synthesized based on the sequence portions in the above-mentioned 3 mutational positions in the nucleotide sequence, it could be analyzed and the above Taq Dye Deoxy<TM> Terminaler Cycle Sequencing Kit (manufactured by Applied Biosystems). It was found that the nucleotide sequences of all four clones were identical to that of SEQ JD NO: 6. Positions 84 and 740 corresponding to SEQ JD NO: 7 were replaced by C and T respectively in these 4 clones. But position 1198 was G in 2 clones and A in other two clones. In other words, it was revealed that there are two types of nucleotide sequences inherent in the original chromosomal DNA. It is not clear whether such differences in the nucleotide sequences derived from homologous chromosomes or from multiple genes. In addition, it has been suggested that the difference in the nucleotide sequence may be due to racial differences, as the poly (A)<+ >RNA procured from Clontech is of a Caucasian origin while the chromosomal DNA is of Japanese origin.

The vector pHGT1 carried by E. coli strain DH5 has been deposited by the present inventors on March 24, 1994 under accession no. FERM BP-4616 at the National Institute of Bioscience and Human-Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

(Example 26)

Expression of human TPO chromosomal DNA in COS 1 cells and confirmation of TPO activity

The plasmid clone pHGT1 obtained by subcloning contained a total of 4 EcoRI recognition sequences, 3 in the inserted group and 1 in the vector. Nucleotide sequence analysis revealed that the entire human TPO protein coding region was included in a DNA fragment of about 4.3 kbp that was taken between the EcoRI recognition sequence closest to the 5' side of the insert and the EcoRI recognition sequence in the vector. This fragment was ligated with an Eco-RI treated expression vector pEF18S and 4 human TPO expression plasmids pEFHGTE No. 1-4 were obtained (see Fig. 11). Purification of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989), thereby obtaining about 250 ng of plasmid DNA.

Transfection into COS 1 cells of the thus obtained clones pEFHGTE no. 1-4 was carried out according to the procedure in example 11. Transfection was carried out with 10 ng of plasmid DNA by the DEAE-dextran method which includes chloroquine treatment. The transfected COS 1 cells were incubated for 3 days at 37°C and then the culture supernatant was collected.

The culture supernatants were thoroughly dialyzed against IMDM culture medium and evaluated by the rat CFU-MK analysis system. TPO activity was detected in a dose-dependent manner in the supernatant of COS 1 cells in which each of the four clones (pEFHGTE no. 1-4) was expressed. In contrast, TPO activity was not detected in the supernatant of COS 1 cells in which the plasmid pEF18S alone was expressed. Representative data with pEFHGTE No. 1 are shown in Fig. 12a. Corresponding results were obtained in the M-07e analysis system. The supernatant of COS 1 cells in which each of the four clones (pEFHGTE No. 1-4) was expressed significantly increased M-07e cell growth in a dose-dependent manner. Representative ways of pEFHGTE No. 1 are shown in FIG. 12b.

These results demonstrated that the plasmid clones pEFHGT nr-1-4 carry functional chromosomal DNA for human TPO.

(Example 27)

Preparation of human TPO deletion strand DNA and its expression in COS 1 cells and confirmation of TPO activity

The results in Example 18 indicate that human TPO can exhibit its biological activity even after removal of its carboxyl third. Thus, to further evaluate the biologically active parts, deletion derivative experiments were performed. In this example, TPO deletion derivatives lacking 20, 40, 60, or 68 amino acids from the carboxyl-terminal end of the TPO protein (amino acids 1 to 231) encoded by pHT 1-231 were examined for their ability to display TPO in vitro biological activity. The shortest derivative (amino acids 1-163) even comprised the amino acid sequences corresponding to TP2/3 of rat plasma TPO described in example 2. The deletion plasmids were produced by PCR using DNA of the plasmid clone pHT1-231 produced in example 18 as template and synthesized oligonucleotides as primers. Sequences of primers used from PCR were as follows:

hTPO-5: 5'-TTT GAA TTC GGC CAG CCA GAC ACC CCG GCC-3'

(SEQ ID NO: 56) (made by adding an EcoRI recognition sequence to the sequence in positions 1 to 21 of SEQ ID NO: 4; this is the identical sequence as described in Fig. 9);

hTPO-S: 5'-TTT GCG GCC GCT CAT TAG CTG GGG ACA GCT GTG GTG GGT-3'

(SEQ ID NO: 57) (an antisense primer corresponding to the sequence at positions 555 to 576 of SEQ ID NO: 4, prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion derivative encoding position 1 to 163 amino acid residues);

hTPO-4:5'-TTT GCG GCC GCT CAT TAC AGT GTG AGG ACT AGA GAG GTT

CTG-3' (SEQ DD NO: 58) (an antisense primer corresponding to the sequence at positions 576 to 600 of SEQ DD NO: 4, prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion derivative encoding for position 1 to 171 amino acid residues);

hTPO-30: 5' TTT GCG GCC GCT CAT TAT CTG GCT GAG GCA GTG AAG TTT

CTG-3' (SEQ DD NO: 59) (an antisense primer corresponding to the sequence at positions 636 to 660 of SEQ DD NO: 4, prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion derivative encoding for position 1 to 191 amino acid residues); and

hTPO-2: 5'-TTT GCG GCC GCT CAT TAC AGA CCA GGA ATC TTG GCT CTG

AAT-3' (SEQ DD NO: 60) (an antisense primer corresponding to the sequence at positions 696 to 720 of SEQ DD NO: 4, prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion derivative encoding for position 1 to 211 amino acid residues).

By using 1 µg of plasmid DNA of the clone pHT 1-231 obtained in example 18 as template, 10 µg of each of the thus synthesized hTPO-5 (for the 5' site) and hTPO-2, -3, - 4 and -S (for the 3' site) as a primer and the GeneAmp<TM><p>cr Reagent Kit with AmpliTaq<TM> DNA Polymerase (manufactured by Takara Shuzo), PCR was performed in a volume of 100 µl using a GeneAmp<TM><p>cr System 9600 (manufactured by PERIQN-ELMER) and repeating a total of 20 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 65°C for 1 minute and synthesis at 72° C for 1 minute, followed by the final incubation at 72°C for 7 minutes. An interesting band thus obtained by PCR was cut with the restriction enzymes EcoRI and Noti and the resulting product subjected to 1% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a major DNA fragment of an expected size amplified by each PCR. The thus isolated DNA fragment was purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and then subcloned into the expression vector pEF18S which had previously been treated with the same restriction enzymes. In this case, Competent-High E. coli DH5 produced by TOYOBO was used as the host strain. From the resulting transformants derived from each PCR, 4 to 5 clones containing the insert of the expected size were selected to prepare plasmid DNA samples essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). In a series of such operations, plasmid DNA samples were obtained from the deletion derivative encoding positions 1 to 163 amino acid residues (pHT1-163 no. 1-5), a deletion derivative encoding position 1 to 171 amino acid residues (pHT1-171 no. 1- 4) a deletion derivative coding for positions 1 to 191 amino acid residues (pHT1-171 no. 1-4) and a deletion derivative coding for positions 1 to 211 amino acid residues (pHT1-211 no. 1-4). The entire length of the amplified regions of each plasmid DNA was subjected to nucleotide sequence analysis to determine its complete identity with the nucleotide sequence shown in SEQ ID NO: 4.

Transfection into COS 1 cells of the thus obtained clones was carried out according to the procedure in example 11. That is, transfection was carried out with 10 (xg of each of the plasmid DNA samples by the DEAE dextran method which includes chloroquine treatment. The transfected COS 1 cells were cultured for 3 days and then the culture supernatant was recovered.

The culture supernatants were thoroughly dialyzed against IMDM culture medium and evaluated by the rat CFU-MK analysis system. The TPO activity was detected in a dose-dependent manner in the supernatant of COS 1 cells in which pHTl-211, pHTl-191, pHTl-171 or pHTl-163 were each expressed. Representative data with pHTl-211 No. 1, pHTl-191 No. 1 and pHTl-171 No. 2 are shown in Fig. 13a and with pHTl-163 no. 2 in fig. 13b. Corresponding results were obtained in the M-07e analysis system. The supernatant from the COSI cells in which pHTl-211, pHTl-191, pHTl-171 or pHTl-163 were each expressed significantly increased M-07e cell proliferation. Representative data with pHTl-211 No. 1, pHTl-191 No. 1, pHTl-171 No. 2 or pHTl-163 No. 2 are shown in Fig. 14.

These results showed that human TPO still retains its in vitro biological activity even after its carboxyl-terminal half past Ser (position 163) is deleted and this strongly suggests that the active biologically active parts of human TPO are located in the amino-terminal half at Ser (position 163).

(Example 28)

Production of C-terminal deletion type human TPO and its expression and activity in COS 1 cells

To analyze the region required for human TPO protein as revealed by its activity, the C-terminal amino acids were further deleted from the deletion derivatives prepared in Example 27 and the TPO activity expressed in the COSI cells was examined. Using the plasmid clone pEF18S-HL34 obtained in example 16, expression plasmids were prepared by deleting nucleotide sequences that code for C-terminal amino acid residues starting from 163 position serine. Construction of these deletion sites was performed using PCR. Sequences of primers prepared for use in PCR are as follows: hTPO-5: 5'-TTT GAA TTC GGC CAG CCA GAC ACC CCG GCC-3' (SEQ ID NO: 61) (prepared by adding an EcoRI recognition sequence to the sequence in positions 1 to 21 of SEQ ID NO: 4; the same sequence is shown in Example 18);

hTPO-150:5'-TTT GCG GCC GCT CAT TAG AGG GTG GAC CCT CCT ACA AGC

AT-3' (SEQ ID NO: 62) (an antisense primer corresponding to the sequence at positions 514 to 537 of SEQ ID NO: 4; prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion mutant which codes for positions 1 to 150 amino acid residues);

hTPO-151: 5'-TTT GCG GCC GCT CAT TAG CAG AGG GTG GAC CCT CCT ACA

A-3<1> (SEQ ID NO: 63) (an antisense primer corresponding to the sequence at positions 518 to 540 of SEQ ID NO: 4; prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the preparation of a deletion mutant encoding positions 1 to 151 amino acid residues);

hTPO-153: <5->TTT GCG GCC GCT CAT TAC CTG ACG CAG AGG GTG GAC CC-3' (SEQ ID NO: 64) (an antisense primer corresponding to the sequence at positions 526 to 546 of SEQ ID NO: 4; prepared by the addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the production of a deletion mutant encoding positions 1 to 153 amino acid residues);

hTPO-154: 5'-TTT GCG GCC GCT CAT TAC CGC CTG ACG CAG AGG GTG GA-3' (SEQ ID NO: 65) (an antisense primer corresponding to the sequence at positions 529 to 549 of SEQ ID NO: 4; prepared by addition of two termination codons TAA and TGA and a Noti recognition sequence for use in the production of a deletion mutant encoding positions 1 to 154 amino acid residues);

hTPO-155; 5'-TTT GCG GCC GCT CAT TAG GCC CGC CTG ACG CAG AGG GT-3' (SEQ ID NO: 66) (an antisense primer corresponding to the sequence at positions 532 to 552 of SEQ DD NO: 4; prepared by addition of two termination codons TAA and TGA and a Noti recognition sequence for use in making a deletion mutant encoding positions 1 to 155 amino acid residues);

hTPO-156: 5'-TTT GCG GCC GCT CAT TAT GGG GCC CGC CTG ACG CAG AG-3' (SEQ DD NO: 67) (an antisense primer corresponding to the sequence at position 535 to 555 of SEQ DD NO: 4; prepared by addition of two termination codons TAA and TGA and a Noti recognition sequence for use in making a deletion mutant encoding positions 1 to 156 amino acid residues); and

hTPO-157: 5'-TTT GCG GCC GCT CAT TAG GGT GGG GCC CGC CTG ACG CA-3' (SEQ DD NO: 68) (an antisense primer corresponding to the sequence at positions 538 to 558 of SEQ DD NO: 4; prepared by addition of two termination codons TAA and TGA and a

Noti recognition sequence for use in the preparation of a deletion mutant encoding positions 1 to 157 amino acid residues).

PCR was performed using 1 µg of plasmid DNA from the clone pEF18S-HL34 prepared in Example 16 as template and 10 µg of each of the thus synthesized oligonucleotides (hTPO-5 for the 5' site and hTPO-150, -151, -153, -154, -155, -156 and -157 for 3'-site) as primers.Using the GeneAmp<TM><p>cr Reagent Kit with AmpliTaq<TM> DNA polymerase (manufactured by Takara Shuzo), The PCR reaction performed in 100 µl volume using the GeneAmp™ PCR System 9600 (manufactured by PERKIN-ELMER) and repeated for a total of 20 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 66°C for 1 minute and synthesis at 72°C for 1 minute followed by the final incubation at 72°C for 7 minutes.A band of interest thus obtained in each PCR was cut with restriction enzymes EcoRI and Noti and resulting product was exposed to 1% agarose gel

(manufactured by FMC BioProducts) electrophoresis to isolate a major DNA fragment of a size expected in each PCR. The thus isolated DNA fragment was purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and then subcloned into the expression vector pEF18S which had been treated beforehand with the same restriction enzymes. In this case, Competent-High E. coli DH5 produced by TOYOBO was used as the host strain. From the resulting transformants promoted for each PCR experiment, 3 to 5 clones containing the insert of the expected size were selected and plasmid DNA samples were prepared essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). In a series of such operations, DNA samples were obtained from a deletion derivative coding for positions 1 to 150 amino acid residues (pHT1-150 no. 21, 22 and 25), a deletion derivative coding for positions 1 to 151 amino acid residues (pHT1-151 no. .16, 17 and 18), a deletion derivative coding for positions 1 to 153 amino acid residues (pHT1-153 Nos. 1 to 5), a deletion derivative coding for positions 1 to 154 amino acid residues (pHT1-154 Nos. 1 to 5), a deletion derivative coding for positions 1 to 155 amino acid residues (pHT1-155 nos. 1 to 5), a deletion derivative coding for positions 1 to 156 amino acid residues (pHTl-156 nos. 1 to 5) and a deletion derivative coding for positions 1 to 157 amino acid residues (pHT1-157 no. 1 to 5).

Of these purified plasmid DNA samples, the sequence of pHTl-150 Nos. 21,22 and 25 and pHTl-151 Nos. 16,17 and 18 was checked against 373A DNA sequences prepared by Applied Biosystems using Taq Dye Deoxy<TM> Terminators Cycle Sequencing Kit (Applied Biosystems), thereby confirming expected TPO cDNA sequences with no substitutions over the entire nucleotide sequence.

Transfection of COS 1 cells with each of the clones thus obtained was carried out according to the procedure of Example 11. That is, transfection was carried out with 10 µg of each of the plasmid samples by the DEAE-dextran method comprising chloroquine treatment, and the culture supernatant was recovered after 3 days of cultivation . The thus obtained culture supernatants were dialyzed against the IMDM culture medium described above and evaluated with the M-07e analysis system. TPO activities were detected in the culture supernatant from COS 1 cells transfected with respective clones encoding C-terminal deletion derivatives comprising 1 to 151 position amino acids, 1 to 153 position amino acids, 1 to 154 position amino acids, 1 to 155 position amino acids, 1 to 156 position amino acids or 1 to 157 position amino acids. All these derivatives contain a cysteine residue at position 151. However, no TPO activity was detected in the culture supernatant of COSI cells transfected with one encoding a C-terminal side deletion derivative comprising 1 to 150 amino acids if the cysteine residue at position 151 was also deleted.

(Example 29)

Preparation of the N-terminal deletion type human TPI and its expression and activity in COS 1 cells

In order to analyze the region necessary for the TPO protein to show its biological activity, the amino acids on one terminal side of the deletion derivatives obtained in Example 28 were further deleted and the expression of TPO activity in the deletion derivatives thus prepared was investigated. Using the plasmid clone pEF18S-HL34 obtained in Example 16 and the plasmid clone pHT1-163 obtained in Example 27, expression plasmids were prepared by deleting the nucleotide sequence encoding N-terminal amino acid residues after the signal sequence. Construction of these deletion plasmids was performed using PCR. Sequences of primers prepared for use in PCR are as follows: hTPO-5: 5' TTT GAA TTC GGC CAG CCA GAC ACC CCG GCC-3" (SEQ DD NO: 69) (prepared by addition of an EcoRI recognition sequence to the sequence in position 1 to 21 of SEQ DD NO: 4; the same sequence shown in Example 18): hTP03:5'-TTT GCG GCC GCT CAT TAT TCG TGT ATC CTG TTC

AGG TAT CC-3' (SEQ DD NO: 70) (an antisense primer corresponding to the sequence at positions 757 to 780 of SEQ DD NO: 4; prepared by addition of two terminal codons TAA and TGA and Noti recognition sequence; the same sequence synthesized for use in the preparation of a deletion derivative coding for positions 1 to 231 amino acid residues);

hTPO-S: 5'-TTT GCG GCC GCT CAT TAG CTG GGG ACA GCT

GTG GTG GGT-3' (SEQ DD NO: 71) (an antisense primer corresponding to the sequence at positions 555 to 576 of SEQ ID NO: 4; prepared from a deletion derivative encoding positions 1 to 163 amino acid residues);

hTPO-13: 5*-AGT AAA CTG CTT CGT GAC TCC CAT GTC CTT

CAC AGC AGA CTG AGC CAG TG-3' (SEQ ID NO: 72) (positions 124 to 173 of SEQ ID NO: 4; prepared for use in the preparation of a derivative in which the amino acid residue in positions 1 to 12 is deleted);

hTPO-13R: 5'-CAT GGG AGT CAC GAA GCA GTT TAC TGG

ACA GCG TTA GCC TTG CAG TTA G-3' (SEQ ID NO: 73) (an antisense primer corresponding to the sequence in positions 64 to 87 and 124 to 148 of SEQ ID NO: 4; prepared for use in the preparation of a derivative in which the amino acid residues in positions 1 to 12 are deleted);

hTPO-7: 5'-TGT GAC CTC CGA GTC CTC AGT AAA CTG CTT

CGT GAC TCC CAT GTC CTT C-3' (SEQ ID NO: 74) (positions 106 to 154 of SEQ ID NO: 4; prepared for use in the preparation of a derivative in which the amino acid residue in positions 1 to 6 has been deleted);

hTPO-7R: 5'-TTT ACT GAG GAC TCG GAG GTC ACA GGA CAG

CGT TAG CCT TGC AGT TAG-3' (SEQ ID NO: 75) (an antisense primer corresponding to the sequence in positions 64 to 87 and 106 to 129 of SEQ ID NO: 4; prepared for use in the preparation of a derivative in which the amino acid residues in position 1 to 6 are deleted);

hTPO-8: 5'-GAC CTC CGA GTC CTC AGT AAA CTG CTT CGT GAC TCC CAT

GTC CTT CAC A-3' (SEQ ID NO; 76) (positions 190 to 157 of SEQ ID NO: 4; prepared for use in the preparation of a derivative in which the amino acid residue in positions 1 to 7 has been deleted); and

hTPO-8R: 5'-CAG TTT ACT GAG GAC TCG GAG GTC GGA CAG

CGT TAG CCT TGC AGT TAG-3' (SEQ DD NO: 77) (an antisense primer corresponding to the sequence in positions 64 to 87 and 109 to 132 in SEQ DD NO: 4; prepared for use in the preparation of a derivative where the amino acid residues in position 1 to 7 are deleted). (1) Preparation of a derivative (pHT13-231) in which the amino acid residue in position 1 to 12 has been deleted.

PCR was carried out using 1.4 µg of plasmid DNA from the clone pEF18S-HL34 prepared in Example 18 as template and 5 µm of each of the thus synthesized oligonucleotides (hTPO-13 and hTP03 in a combination and hTPO-5 and hTPO -13R in another combination) as primers. Using the GeneAmp<TM><p>cr Reagent Kit with AmpliTaq<TM> DNA Polymerase (manufactured by Takara Shuzo), the PCR reaction was performed in 100 µl volume using the GeneAmp<TM> PCR System 9600 (manufactured by PERKIN - ELMER) and, after 5 minutes of denaturation at 95°C, a total of 30 cycles were repeated, each cycle comprising denaturation at 95°C for 1 minute, annealing at 65°C for 1 minute and synthesis at 72°C for 1 minute , followed by the final incubation at 72°C for 7 minutes, was carried out. Each of the PCR products was subjected to 1.2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a major DNA fragment of an expected size which was then purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad). and dissolved in 15 ul TE buffer. Subsequently, the second PCR was carried out using a 1 µl portion of each of the thus prepared solutions as a template. The second PCR was performed using 5 µm of each of the synthesized primers (hTPO-5 and hTP03). Using the GeneAmp™ PCR Reagent Kit with AmpliTaq™ DNA polymerase (manufactured by Takara Shuzo), the PCR reaction was performed in 100 µl volume using the GeneAmp™ PCR System 9600 (manufactured by PERKIN-ELMER) and, after 5 minutes of denaturation at 95°C, repeating a total of 30 cycles, each cycle consisting of denaturation at 96°C for 1 minute, annealing at 60°C for 1 minute and synthesis at 72°C for 1 minute, followed by the final incubation at 72°C for 7 minutes. Each of the PCR products was exposed to 1.2% agarose gel (manufactured by FMC BioProducts). electrophoresis to isolate a major DNA fragment of an expected size which was then purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and dissolved in 15 µl TE buffer. After cutting with restriction enzymes EcoRI and Noti, the resulting solution was subjected to extraction with the same volume of phenol/chloroform and subsequent ethanol precipitation. After centrifugation, the resulting pellet was dissolved in 15 µl TE buffer and then subcloned into the expression vector pEF18S which had been pre-treated with the same restriction enzymes. In this case, Competent-High E. coli DH5 manufactured by TOYOBO was used as the host strain. From resulting transformants, 45 clones containing the insert at the expected size were selected to prepare plasmid DNA samples essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor laboratory Press, 1989). In this way, it is possible to obtain plasmid DNA of a deletion derivative (pHT13-231) which codes for a protein molecule in which the amino acid residue in position 1 to 12 has been deleted from the original amino acid residue in position 1 to 231. When these 45 clones are exposed to PCR using primers hTPO-5 and hTP03, inserts of approximately the expected size were confirmed in 8 clones. Of these, the sequence of 3 clones was checked by 373A DNA sequences produced by Applied Biosystems using the Taq Dye Deoxy<TM> Terminaler Cycle Sequencing Kit (Applied Biosystems) thereby obtaining a clone pHT13-231 No. 3 with a TPO cDNA sequence was designed and a deletion as expected with no substitutions and the like over the entire nucleotide sequence. (2) Preparation of a derivative (pHT7-163) in which the amino acid residue in positions 1 to 6 is deleted.

In this case, the amino acid at position 163 was used as the C-terminal for the preparation of deletion derivatives, although the amino acid at position 231 was shown as a C-terminal end of the TPO protein for the preparation of deletion derivatives in step 1 above, it was found that TPO the activity could be expressed even when the C-terminal amino acids were further deleted. On the basis of the same reasoning, the amino acid in position 163 was used as C-terminal also in the subsequent step (3). PCR was performed using 1.4 ng plasmid DNA of clone pHT1-163 obtained in example 27 as template and 5 µm of each of the synthesized oligonucleotides (hTPO-7 and hTPO-S in a combination and hTPO-5 and hTPO-7R in a different combination) as primers. Using the GeneAmp<TM><p>cr Reagent Kit with AmpliTaq™^ DNA polymerase (manufactured by Takara Shuzo), the PCR reaction was performed in 100 (µl volume using GeneAmp™^ < p>cr System 9600 (manufactured by PERKIN-ELMER) and, after 5 minutes of denaturation at 95°C, repetition of a total of 30 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 65°C for 1 minute at synthesis for 72 minutes, followed by the final incubation at 72°C for 7 minutes.Each of the PCR products was subjected to 1.2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate the main DNA fragment of an expected size which is then purified by using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and dissolved in 15 µl TE buffer. dishes, a second PCR was performed using 1 µl aliquot of each prepared solution as a template.

The second PCR was performed using 5 µm of each of the synthesized primers (hTPO-5 and hTPO-S). Using the GeneAmp™ PCR Reagent Kit with AmpliTaq<TN>DNA polymerase (manufactured by Takara Shuzo), The PCR reaction performed in 100 µl volume using the GeneAmp™ PCR System 9600 (manufactured by PERKIN-ELMER) and, after 5 min denaturation at 95°C, repetition of a total of 30 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 60°C for 1 minute and synthesis at 72°C for 1 minute, followed by the final incubation at 72°C for 7 minutes.Each of the PCR products was subjected to 1.2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a major DNA fragment of an expected size which was then purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad) and dissolved in 15 µl TE buffer. After cutting with restriction enzymes EcoRI and Noti, the resulting solution subjected to extraction with the same volume of phenol/chloroform and then ethanol field. After centrifugation, the resulting precipitate was dissolved in 15 µl TE buffer which was then subcloned into the expression vector pEF18S which had been treated beforehand with the same restriction enzymes. In this case, Competent-High E. coli DH5 produced by TOYOBO was used as the host strain. From the resulting transformants, 30 clones containing inserts of the expected size, selected to prepare plasmid DNA samples essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring HARBOR laboratory Press, 1989).In this way it was possible to obtain plasmid DNA of a deletion derivative (pHT7-163) which codes for a protein molecule where the amino acid residue in positions 1 to 6 had been deleted from the original amino acid residues from position 1 to 163 in SEQ ID NO: 4. When these were subjected to PCR using primers hTPO-5 and hTPO-S, inserts of approximately the expected size were confirmed in all clones, of which the sequence of 3 clones was checked at 373A DNA synthesis prepared by Applied Biosystems using the Taq Dye Deoxy<TM> Terminator Cycle Sequencing Kit (Applied Biosystems), thereby obtaining 2 clones, pHT7-163 No. 4 and No. 29, each with a TPO cDNA sequence as indicated and a deletion as expected without substitution and the like over the entire nucleotide sequence. (3) Preparation of a derivative (pHT8-163) in which the amino acid residue in positions 1 to 7 has been deleted A derivative (pHT8-163) by deletion of amino acid residues in position 1 to 7 was prepared essentially in the same way as in the above case for the preparation of a derivative (pHT7-163) with deletion of the amino acid residue in positions 1 to 6. PCR was performed using 1.4 µg of plasmid DNA and the clone pHT 1-163 prepared in Example 27 as a template and 5 nm of each of the synthesized oligonucleotides (hTPO-8 and hTPO-S) in one combination and hTPO-5 and hTPO-8R in another combination) as primers. Using the GeneAmp<TM><p>QR Reagent Kit with AmpliTaq<TM> DNA polymerase (manufactured by Takara Shuzo), the PCR reaction was performed in 100 µl volume using the GeneAmp™ PCR System 9600 (manufactured by PERKJN-ELMER) and, after 5 minutes of denaturation at 95°C, repetition of a total of 3o cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 65°C for 1 minute and synthesis at 72°C for 1 minute, followed by a final incubation at 72°C for 7 minutes. Each of the PCR products was subjected to 1.2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate a major DNA fragment of an expected size which is then purified using the Prep-A-Gene DNA Purification Kit (manufactured by BioRad) and dissolved in 15 µl TE buffer. Then the second PCR was carried out using 1 µl portion of each of the thus prepared solutions of a template.

The second PCR was performed using 5 µm of each of the synthesized primers (hTPO-5 and hTPO-S). Using the GeneAmp™ PCR Reagent Kit with AmpliTaq™ DNA Polymerase (manufactured by Takara Shuzo), the PCR reaction was performed in 100 µl volume using the GeneAmp™ PCR System 9600 (manufactured by PERKIN-ELMER) and, after 5 minutes of denaturation at 95°C was repeated for a total of 30 cycles, each cycle comprising denaturation at 95°C for 1 minute, annealing at 60°C for 1 minute and synthesis at 72°C for 1 minute, followed by the final incubation at 72°C for 7 minutes. Each of the PCR products was subjected to 1.2% agarose gel (manufactured by FMC BioProducts) electrophoresis to isolate the main DNA fragment of expected size which was then purified using the Prep-A-Gene DNA Purification Kit (manufactured by BioRad). and dissolved in 15 (µl TE buffer. After cutting with restriction enzymes EcoRI and Noti, the resulting solution was subjected to extraction with the same volume of phenol/chloroform and then ethanol field. After centrifugation, the resulting precipitate was dissolved in 15 (µl TE buffer so subclone into the expression vector pEF18S which had been treated beforehand with the same restriction enzymes. In this case, Competent-High E. coli DH5 produced by TPYOBO was used as the host strain. From the resulting transformants, 30 clones containing the insert of the expected size were selected to prepare plasmid DNA samples essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989). In this way it was possible to obtain plasmid DNA of a deletion derivative (pHT8-163) which codes for a protein molecule in which the amino acids in position 1 to 7 have been deleted from the original amino acid residues in position 1 to 163 in SEQ DD NO: 4. When these clones were subjected to PCR using primers hTPO-5 and hTPO-S, inserts of approximately the expected size were confirmed in all clones. Of these, the sequence of 3 clones was checked with 373A DNA sequences produced by Applied Biosystems using the Taq Dye Deoxy<TM> Terminator Cycle Sequencing Kit (Applied Biosystems), thereby obtaining 2 clones, pHT8-163 no. 33 and no. .48, each with a TPO cDNA sequence as indicated and a deletion as expected with no substitutions and the like over the entire nucleotide sequence.

Expression of deletion derivatives in COS 1 cells and confirmation of TPO activity Transfection of COS 1 cells with each of the thus obtained deletion clones was carried out according to the procedure in Example 11. That is, the transfection was carried out with 10 (µg of each of the plasmid DNA samples at The DEAE-dextran method comprising chloroquine treatment and the culture supernatants were recovered after 3 days of cultivation.The thus obtained culture supernatants were thoroughly dialyzed against IMDM culture medium described above and evaluated with the M-07e analysis system.

As a result, weak but significant TPO activity was detected in the culture supernatant of COS 1 cells transfected with a clone encoding the deletion derivative consisting of 7 to 163 position amino acids. However, no TPO activity was detected in the culture supernatant from COS 1 cells transfected with clones encoding the deletion derivatives consisting of the amino acids in position 8 to 163 or of the amino acids in position 13 to 231.

(Example 30)

Preparation of complete amount of human TPO cDNA plasmid (pHTP1)

An expression vector with all amino acid coding regions from human TPO cDNA according to SEQ ID NO: 6 was constructed for expression in mammalian cells.

PCR was performed as follows to prepare a DNA fragment covering all human TPO cDNA coding regions.

The nucleotide sequence used as a primer is as follows:

hTPO-I: 5'-TTG TGA CCT CCG AGT CCT CAG-3' (SEQ ID NO: 78)

(positions 105 to 125 of SEQ ID NO: 6);

SA:5'-CAG GTA TCC GGG GAT TTG GTC-3' (SEQ ID NO: 79) (an antisense primer corresponding to the sequence at positions 745 to 765 of SEQ ID NO: 6);

hTPO-P: 5'-TGC GTT TCC TGA TGC TTG TAG-3' (SEQ ID NO:80)

(positions 503 to 523 of SEQ ID NO: 6); and

hTPO-KO: 5<*->GAG AGA GCG GCC GCT TAC CCT TCC TGA GAC

AGA TT-3' (SEQ ID NO: 81) (a sequence prepared by adding a recognition sequence for restriction enzyme Noti and a GAGAGA sequence (SEQ ID NO: 82) to give an antisense sequence corresponding to the sequence at positions 1066 to 1086 of

SEQ ID NO: 6).

The first PCR was performed using 300 ng of the clone pEF18S-HL34 prepared in Example 16 as template. Using 0.5 [mu]m of each of the primers hTPO-I and SA and 1 unit of Vent R<TM> DNA polymerase (manufactured by New England BioLabs), PCR (repetition of a total of 30 cycles, each cycle including incubation at 96° C for 1 minute, at 62°C for 1 minute and 72°C for 1 minute, followed by a final incubation at 72°C for 7 minutes) were performed. The composition of the reaction solution in final concentration is as follows: 10 mM KC1, 10 mM (NH2 SO2, 20 mM Tris-HCl (pH 8.8), 2 mM MgSO4, 0.1% Triton X-100 and 200 [mu]m dNTP mixture.

A microgram portion of a commercially available poly(A)<+> RNA preparation derived from a human normal liver (manufactured by Clontech) was heated to 70°C for 10 minutes, rapidly cooled on ice, and then mixed with 10 mM DTT, 500 [ im dNTP mixture, 25 ng random primer (manufactured by Takara Shuzo), 10 units of RNase inhibitor (manufactured by Boehringer-Mannheim) and 200 units of SuperScript^<M> jj RNase H" (manufactured by LIFE TECHNOLOGIES), followed by 1 hour incubation at 37°C to carry out the synthesis of cDNA. Then, a second PCR (repetition of a total of 30 cycles, each cycle comprising incubation at 96°C for 1 minute, at 58°C for 1 minute and at 72°C for 1 minute) was performed using 1/20 volume of the reaction solution of synthesized cDNA as template and 2.5 µm of each of the primers hTPO-P and hTPO-KO and 2.5 units of AmpliTaq<TM> DNA polymerase (manufactured by Takara Shuzo).

Each of the resulting solutions of the first and second PCR was subjected to 1% agarose gel electrophoresis to isolate receptor-specific main DNA fragments of the expected size which are then purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad). . PCR was then carried out using 1/20 volume of each of the thus prepared solutions as a template. This reaction (heating at 96°C for 2 minutes followed by repetition of a total of 3 cycles, each cycle comprising incubation at 96°C for 2 minutes and 72°C for 1 minute and subsequent incubation at 72°C for 7 minutes) was performed using one unit of Vent R<TM> DNA polymerase (manufactured by New England BioLabs). The resulting reaction solution was mixed with 1 Hm of each of hTPO-I and hTPO-KO, heated at 96°C for 2 minutes, subjected to 25 cycles of reaction, each cycle comprising incubation at 96°C for 1 minute, 62°C for 1 minute and 72°C for 1 minute and then incubation at 72°C for 7 minutes. The resulting reaction solution was extracted with the same volume of water-saturated phenol-chloroform and then with the same volume of chloroform, the extract was subjected to ethanol precipitation (2.5 volumes of ethanol in the presence of 0.3 M sodium acetate and 0.5 µl of glycogen prepared by Boehringer-Mannheim) to recover DNA. The DNA thus recovered was cut with restriction enzymes BamHI and Noti and subjected to 1% agarose gel electrophoresis and a major band of an expected size isolated thus, was purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad), ligated with pBluescript II SK<+> vector (manufactured by Stratagene) which had been cut beforehand with the restriction enzymes BamHI and Noti and then transformed into Competent-High E. coli DH5 (manufactured by TOYOBO). From the resulting colonies, 4 clones were selected to prepare plasmid DNA samples. The sequence of purified plasmid DNA samples was checked with 373A DNA sequences produced by Applied Biosystems using the Taq Dye Deoxy<TM>Terminaler Cycle Sequencing Kit (Applied Biosystems), thereby obtaining a clone, pBLTP with a TPO cDNA sequence designed without substitution in the nucleotide sequence in the region between BamHI and Noti.

The clone pBLTP was cut with restriction enzymes EcoRI and BamHI and subjected to 1% agarose gel electrophoresis and a high molecular weight band thus isolated was purified using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad). In the same way, pEF18S-HL34 was treated with the restriction enzymes to purify a DNA fragment around 450 bp. Each of the DNA samples thus obtained was subjected to ligation and transformed into Competent-High E. coli DH5 (manufactured by TOYOBO). Plasmid DNA samples were prepared from the resulting colonies to obtain a clone, pBLTEN, containing the human TPO cDNA insert. The thus obtained pBLTEN was cut with restriction enzymes EcoRI and Noti and subjected to 1% agarose gel electrophoresis and a DNA fragment of about 1200 bp was thus isolated using the Prep-A-Gene DNA Purification Kit (manufactured by Bio-Rad), ligated with the expression vector pEF18S which had been cut beforehand with the same restriction enzymes and then transformed into Competent-High E. coli DH5 (manufactured by TPYOBO). Plasmid DNA samples were

prepared from the resulting colonies to obtain the interesting clone pHTP1, which contains the entire human TPO cDNA coding region. Plasmid DNA was prepared from this clone in large quantity and used in the following experiments. Preparation of plasmid DNA was performed essentially according to the procedure described in Molecular Cloning (Sambrook et al., Cold Spring Harbor Laboratory Press, 1989).

(Example 31)

Contraction of mammalian expression plasmid. pDEF202- hTPO- Pl. for expression of human TPO in Chinese hamster ovary (CHO) cells

One microgram of plasmid pMG1 containing mouse dihydrofolate reductase (DHFR)

minigene was cut with restriction enzymes EcoRI and BamHi and subjected to agarose gel electrophoresis to isolate a fragment (about 2.5 kb) containing mouse DHFR

the minigene. The isolated DNA fragment was dissolved in 25 ml of a reaction solution composed of 50 mM Tris-HCl (pH 7.5), 7 mM MgCl2, 1 mM 2-mercaptoethanol and 0.2 mM dNTP mixture, mixed with 2 units of Klenow's fragment and then incubated at room temperature for 30 minutes to form blunt ends at both ends of the DNA

the fragment. After phenol/chloroform treatment and ethanol precipitation, the resulting fragment was dissolved in 10 ml of a TE solution (10 mM Tris-HCl (pH 8.0)/l mM

EDTA). The mammalian expression vector pEG18S was cut with restriction

tion enzyme Smal, dephosphorylated with alkaline phosphatase (prepared by Takara Shuzo) and ligated with a DNA fragment containing the mouse DHFR minigene using T4

DNA ligase (prepared by Takara Shuzo) to obtain an expression vector pDEF202.

Then the constructed expression vector pDEF202 was cut with

restriction enzymes EcoRI and Spel, and then subjected to agarose gel electrophoresis to isolate a larger DNA fragment. Human TPO cDNA which had been obtained by cutting plasmid pHTP1 containing human TPO cDNA (clone pl) when restriction enzymes EcoRI and SpeI was ligated with this linearized pDEF202 vector using T4 ligase (manufactured by Takara Shuzo) to obtain a plasmid pDEF202 - hTPO-P1 for expression of human TPO cDNA. This engineered plasmid contains SV40

origin of replication, human elongation factor 1-a promoter and SV40 early polyadenylation signal for transcription of human TPO cDNA, mouse DHFR minigene and pUC18 origin of replication and fi-lactamase gene (Amp<1>).

(Example 32)

Expression of human TPO in CHO cells

CHO cells (a DHFR strain; Urlaub and Chasein, Proe. NAtl. Acad. Sei. USA, vol. 77 (p4216,1980) were maintained in α-minimum essential medium (α-MEM(-), supplemented with hypoxanthine and thymidine ) containing 10% fetal calf serum (FCS) using 6 cm diameter tissue culture dishes (by Falcon). The transformation of CHO cells with pDEF202-hTPO-P1 plasmid was performed by the calcium phosphate method (CellPhect, manufactured by Pharmacia) as described below. Ti micrograms of the plasmid pDEF202-hTPO-P1 prepared in Example 31 was mixed with 120 ml of buffer A and 120 ml of H 2 O and the resulting mixture was incubated at room temperature for 10 minutes.The solution was further mixed with 120 ml of buffer B and placed at room temperature in another 30 minutes. This DNA mixing solution was then added to the culture incubated for 6 hours in a CO2 incubator. After 6 hours, the culture was washed twice with serum-free oc-MEM(-), treated with a-MEM(-) containing 10% dimethyl sulfoxide for 2 minutes, and then incubated in 2 days in non-selection medium) (α-MEM(-) supplemented with hypoxanthine and thymidine) containing 10% dialysis FCS. After 2 days of cultivation, the cells were selected in a selection medium (oc-MEM(-)) containing 10% dialysis FCS. For selection of transformed cells with a DHFR-positive phenotype, the cells were treated with trypsin/EDTA solution and resuspended with a selection medium. The cells in 6 cm tissue culture plates were divided into 5 10 cm tissue culture plates or 12 24 well tissue culture plates and then cultured in the selection medium. The culture medium was changed at two-day intervals. Human TPO activity in the cultured medium of CHO cells with a DHFR-positive phenotype was measured by CFU-MK analysis, M-07e analysis or Ba/F3 analysis. The cells secreting human TPO in the culture medium were further selected in the selection medium supplemented with 25 nM methotrexate to isolate the cell clone that produced a greater amount of human

TPO.

In this case, transfection of CHO cells can also be performed by performing cotransfection of CHO cells with pHTPl and pMGl plasmids.

CHO strain (CHO-DUKXB11) transfected with plasmid pDEF202-hTPO-Pl has been deposited by the present applicants on January 31, 1995 under accession number FERM BP-4988, at the National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

(Example 33)

Construction of a recombinant vector BMCHSneo-hTPO-P1 for use in X 63. 6. 5. 3. cells

One microgram of a mammalian expression vector pBMCGSneo was cut with restriction enzymes XhoI and Noti and subjected to agarose gel electrophoresis to isolate the vector DNA portion. This DNA fragment and human TPO cDNA (P1 clone) which had been obtained by cutting the plasmid pBLTEN containing human TPO cDNA with restriction enzymes XhoI and Noti, were ligated using T4 ligase to obtain the expression vector of interest, pBMCGSneo-hTPO-P1. This expression plasmid contains cytomegalovirus early promoter, parts of rabbit B-globin gene derived intron and polyadenylated region for transcription of human TPO cDNA, human B globin gene, 69% of bovine papillomavirus 1 gene, thymidine in kinase promoter and polyadenylation region, phosphotransferase I gene (neomycin resistant gene) and pBR322 origin of replication and 8-lactamase gene (Amp<1>), human TPO cDNA was ligated at a downstream site of cytomegalovirus promoter.

(Example 34)

Expression in X 63. 6. 5. 3. cells

X 63.6.5.3. cells were grown in Dulbecco's minimum essential medium (DME) containing 10% fetal bovine serum. Transfection of X63.6.5.3. cells with the BMCGSneo hTPO plasmid was performed by the electrophoresis method described below. 20 micrograms of the plasmid BMCGSneo-hTPO-P1 prepared in Example 33 was added to 750 µl of the DME medium containing 10 3 cells and the mixture was placed in an electroporation cuvette with a gap of 4 mm and left for 10 minutes at 4°C. The envelope was then placed in an electroporation apparatus (BTX 600, manufactured by BTX) to perform gene transfer under conditions of 380 V, 25 mF and 24 angstroms. After gene transfer, the cuvette was left for 15 minutes at 4°C and the cells were then washed with DEM containing 10% fetal bovine serum. The transfected cells were suspended in 50 ml of DEM containing 10% fetal bovine serum, suspended in wells of 5 96-well culture plates and then cultured for 2 days in a CO2 incubator. After 2 days of cultivation, the culture medium was changed with DMEM containing 10% fetal bovine serum and 1 mg/ml G418 (manufactured by GD3SCO) and medium changes were performed every 3 days thereafter. The human TPO activity in the culture medium of transformants was measured with the CFU-MK, M-07e or Ba/F3 assay. The transformants secreting human TPO in the culture medium were cloned twice at limiting dilution to establish a human TPO producing cell line.

(Example 35)

Large-scale expression of human TPO in COS 1 cells

Transfection of COS 1 cells was performed according to the DEAE-dextran method comprising chloroquine treatment as described in Example 11. COS 1 cells (ATCC CRL1650) were grown in IMDM containing 10% (v/v) FCS for use on a collagen coated 175 cm< 2> culture flask at 37°C in a 5% carbon dioxide incubator until the cells became approximately 100% confluent. To perform collagen coating of the culture flask, 25 ml of a collagen solution (Cellmatrix type I-C, manufactured by Iwaki) whose concentration had been adjusted to 0.3 mg/ml with 1 mM HC1 was added to each 175 cm<2> culture flask, the collagen solution was recovered after 1 hour by standing at room temperature and then the bottles thus treated were washed once with 20 to 50 [mu]l of PBS. Transfection was performed by mixing 20 ml IMDM solution containing 250 µg/ml DEAE-dextran (manufactured by Pharmacia), 60 µm chloroquine (manufactured by Sigma) and 10% (v/v) Nu serum (manufactured by Collaborative) with plasmid pHTPl (40 µg) that had been dissolved in 500 µl HBS, adding the resulting mixture to the COS 1 cells described above contained in a 175 cm<2> culture flask that had been washed once with IMDM immediately before transfection, and then washing the cells for 3 hours at 37°C in a 5% carbon dioxide incubator. Then the culture supernatant was removed by suction and the cells were washed once with IMDM, mixed with 50 ml of a serum-free medium and then cultured for 5 days at 37°C in a 5% carbon dioxide incubator to recover the culture supernatant. In one operation, 100 to 260 175 cm<2> culture flasks were used and 5 to 13 liters of the culture supernatant were recovered. The serum-free medium was prepared by adding IMDM with 5 µg/ml insulin (manufactured by Sigma), 5 µg/ml transferrin (manufactured by Sigma), 10 µm monoethanolamine (manufactured by Wako Pure Chemical Industries), 25 nM sodium selenite (manufactured by Sigma) and 200 µg/ml BSA (an acid-free highly purified bovine albumin, manufactured by Nichirei).

(Example 36)

Purification of human full-length TPO (derived from expression plasmid pHTPl.Pl clone) expressed in COS 1 cells and its biological activity

Examples of the activity and purification of human full-length TPO (derived from expression plasmid pHTP1) expressed in COS 1 cells are described below. A partially purified product was first prepared to investigate the platelet increasing activity and then the purification was carried out to obtain a highly pure product. The M-07e analysis system was mainly used for the measurement of TPO activity in the following manufacturing steps. TPO activity was likewise found in the rat CFU-MK assay system. In performing these assays, human serum albumin (HSA) was added to each sample to a final concentration of 0.02 to 0.05%.

First, COS 1 cells were transfected with the plasmid pHTP1 according to the method of Ohashi and Sudo (Hideya Ohashi and Tadashi Sudo, Biosci. Biotech. Biochem., 58 (4),

758-759,1994) were cultured for 5 days at 37°C in 5% CO2 using a serum-free IMDM culture medium containing 0.2 g BSA, 5 mg bovine serum insulin, 5 mg human transferrin, 0.02 mM monoethanolamine and 25 nM sodium selenite, in a 1000 ml, thereby obtaining around 7 liters of serum-free culture supernatant. To the serum-free culture supernatant was added the proteolytic enzyme inhibitors p-APMSF and Pefabloc SC (4-(2aminoethyl)-benzenesulfonyl fluoride hydrochloride, manufactured by Merk, catalog no. 24839) to a final concentration of about 1 mM, followed by filtration using a 0.2 ( xm filter to recover the supernatant This was concentrated by a factor of about 10 using an ultrafiltration unit (Omega Ultrasette, nominal molecular weight cutoff is 8,000, manufactured by Filtron; or PLGC Pellicon Cassette, nominal molecular weight cutoff is 10,000, manufactured by Millipore) . obtain 804 ml of the solution containing 1.5 M final concentration of ammonium sulfate and then the solution thus prepared was placed at a flow rate of 10 ml/min on a M acro-Prep Methyl HIC column (5 cm in diameter and 9 cm in height, manufactured by Bio-Rad, catalog no. 156-0080) which had been equilibrated beforehand with 20 mM sodium citrate buffer (pH 5.2) containing 1.5 M ammonium sulfate. After loading the sample, elution was performed with a 20 mM sodium citrate buffer (pH 5.2) containing 1.25 M sodium sulfate to obtain direct through fraction Fl (2.384 ml; protein concentration, 0.864 mg/ml; total protein, 2.061 mg; and relative activity, 6,000).

Then, the elution solution was changed to 20 mM Na citrate, pH 5.8, to collect fraction F2 (1.092 ml; protein concentration, 0.766 mg/ml; total protein, 847 mg; and relative activity 150,000).

The thus obtained Macro-Prep Methyl HIC column fraction F2 (1,081 ml) was put at a flow rate of 10 ml/minute on an SP Sepharose Fast Flow column (3 cm diameter and 10 cm length, manufactured by Pharmacia Biotech, catalog no. 17- 0729-01) which had been equilibrated beforehand with 20 mM sodium citrate buffer (pH 5.8). After loading the sample, the elution was performed with 20 mM sodium citrate buffer (pH 5.6) containing 110 mM NaCl to obtain fraction Fl (2.266 ml, protein concentration 0.270 mg/ml; total protein 610 mg; and relative activity 30,000). Then, the elution solution was changed to 20 mM sodium citrate buffer (pH 5.4) containing 400 mM NaCl to collect fraction F2 (856 ml; protein concentration, 0.189 mg/ml; total protein, 162 mg; and relative activity 300,000). Then, the elution solution was changed to 20 mM sodium citrate buffer (pH 5.2) containing 1000 mM NaCl to collect fraction F3 (370 ml; protein concentration 0.034 mg/ml; total protein 12.6 mg; and relative activity 150,000).

Main TPO active fraction F2 (845 ml) from the SP Sepharose Fast Flow column step was mixed with about 10% final concentration of 1-propanol and set at a flow rate of 3 ml/min on LA-WGA column (2 cm diameter and 14 cm height) produced by Honen, catalog no. WG-007) which had been equilibrated beforehand with 20 mM sodium citrate buffer (pH 5.4) containing 400 mM NaCl. After loading the sample, elution was performed using a mixture of 20 mM sodium citrate buffer (pH 5.4) containing 400 mM NaCl with 1-propanol (9:1) to obtain fraction Fl

(64.4 ml; protein concentration, 0.0178 mg/ml; total protein, 1.15 mg and relative activity 17.220). Then, the elution solution was changed to 20 mM sodium citrate buffer (pH 6.1) containing 0.4 M GlcNAc and 10% 1-propanol to collect eluted fraction F2 (45 ml; protein concentration, 0.0104 mg/ml; total protein, 0.470 mg; and relative activity 675,000) .

Main TPO active fraction F2 (340 ml) from the LA-WGA column operation was mixed with about 0.005% in final concentration of TFA and subjected to YMC-Pack CN-AP (6 mm in diameter and 250 mm in height, manufactured by YMC, (catalog no. AP-513) column chromatography. That is, using 0.1% TFA as eluent A and 1-propanol containing 0.05% TFA as eluent B, the sample was loaded at a flow rate of 0.6 ml/min on the EMC-Pack The CN-AP column which had previously been equilibrated with 15% B. After loading the sample, the propanol concentration was increased from 15% B to 25% B and the elution was carried out with a linear gradient from 25% to 50% B during 65 minutes while collecting the eluates in portions of 1.5 ml (2.5 min.) portions in polypropylene tubes.

A 0.5 µl portion (1/3,000 fraction) of the eluate in each tube was mixed with HSA concentrated by ultrafiltration to obtain 0.25 ml of IMDM assay culture solution containing 0.05% HSA for use in identification of TPO active fractions in the assay. As a result, strong TPO activity, (relative activity of about 630,000 to 4,800,000) was found in tubes No. 24 to 30 (within the range of 35.0 and 42.0% as propanol concentration) which was collected as TPO active fraction FA (13.5 ml).

The fraction FA obtained by the YMC-Pack CN-AP column step was subjected to Capcell Pak Cl 300A (4.6 mm diameter and 150 mm height, manufactured by Shiseido, catalog no. CI TYPE:SG300A) column chromatography using 0.1% TFA as eluent A and 1-propanol containing 0.05% TFA as eluent B. A portion (8.9 ml) of fraction FA obtained by YMC-Pack CN-AP column operation was mixed with 0.3 ml glycerol, concentrated by centrifugation evaporation and then mixed with about 2.5 ml of 10% B, and the resulting solution was then applied at a flow rate of 0.4 ml/min. on the Capcell Pak Cl 300A column that had been equilibrated beforehand with 20% B. After single loading, the elution was performed first with 20% B for 5 minutes as a linear gradient from 20% B to 40% B for 50 minutes while collecting the eluates in 1 ml (2.5 minute) portions in polypropylene tubes.

A 1 ni portion (1/1,000 fraction) of the eluate in each tube was mixed with HSA concentrated ultrafiltration to obtain 0.25 ml IMDM assay culture solution containing 0.05% HSA for use in identification of TPO active fractions in the assay. As a result, strong TPO activity (relative activity of about 3,000,000 to 22,500,000) was found in tubes No. 20 to 23 (within the range of 28.5 and 32.5% as propanol concentration) which was collected as a high activity fraction.

(Example 37)

Molecular weight measurement of human TPO expressed in COS 1 cells

It was assumed that the molecular weight of human complete full-length TPO (derived from the expression plasmid pHTP1, F1 clone) expressed in COS 1 cells would be greater than the molecular weight and can be assumed from the size of the peptide chain, due to the addition of sugar chain. Consequently, as a first step, a partially purified TPO fraction was prepared as follows from a culture supernatant containing the human full-length TPO (derived from expression plasmid pHTP1, F1 clone) expressed in COS 1 cells. That is, using eluent A (0.1% trifluoroacetic acid (TFA)) and eluent B (1-propanol containing 0.05% TFA), a 0.3 ml portion of the culture supernatant was placed on YMC-Pack PROTEIN-RP (0.46 mm in diameter and 15 cm height, manufactured by YMC, Catalog No. A-PRRP-33-46-25) column that had been previously equilibrated with 25% B, 25% B was passed through the column for 5 minutes at a flow rate of 0, 4 ml/min. and fractionation was performed with linear density gradient from 25% B to 50% B over 50 minutes. TPO activity was found in eluates in the range from 34.5% to 43.5% propanol concentration and these eluates were evaporated to dryness by centrifugation evaporation to obtain a partially purified sample.

Subsequently, the TPO activity was examined by the M-07e analysis system after extraction of protein from a gel by SDS-PAGE electrophoresis carried out under non-reducing conditions as described in example 1, or molecular weight was measured based on production-treated DPCHT. markers using ay a western analysis which will be described later in example 45. As a result, TPO activity was found in a wide apparent molecular weight range of about 69,000 to 94,000, thus confirming the variation in molecular weight. Apparent molecular weight of TPO based on reduction-treated DPCJU markers after its N-glycosidic linkage type sugar chain cleavage with N-glycanase (manufactured by Genzyme, catalog no. 1472-00) was thus measured to be 36,000 to 40,000, which is greater than the molecular weight of about 3S .000 which is estimated from the size of the peptide chain, which indicates the presence of also an O-glycosidic bond type sugar chain.

Likewise, the apparent molecular weight of the human full-length TPO (derived from expression plasmid pHTP1, P1 clone) expressed in COS 1 cells was found to be 63,000 to 83,000.

(Example 38)

Biological characteristics of human TPO

Effects of human TPO on human and rat hematopoietic cells were investigated, mainly using culture supernatants from COS 1 cells transfected with pHTF1.

In the colony assay using the CD34<+>DR<+> cell fraction prepared from human cord blood, human TPO stimulated the formation of significant numbers of megakaryocyte colonies. For example, 11.5 megakaryocyte colonies were formed from 6000 CD34<+>DR<+> cells in the presence of 10% (v/v) culture supernatant from COS 1 cells transfected with pHTl-231. To examine the effect of human TPO on human peripheral blood megakaryocyte stem cells, CD34<+> cells were prepared from human peripheral blood as follows. A leukocyte fraction obtained from human peripheral blood by a Ficoll-Paque density gradient was depleted of cells adhering to plastic dishes. The non-adherent cells were then depleted of SB A (soybean agglutinin) positive cells by panning using the AIS Microselecter SBA (Asahi Medical). The resulting cells were incubated on the AIS Microselecter CD34 and the adsorbed cells, CD 34<+> cells, were collected when the CD3<+ >cells were cultured for 10 days in the presence of culture supernatant from transfected COS 1 cells in a liquid medium, selective proliferation of large size Gpnb/DIa<+> cells were observed and increase in ploidy was found in these GpIlb/IIIa<+> cells. These results strongly suggest that human TPO exerts its selective action on human megakaryocyte lineages and stimulates their proliferation and differentiation.

In the colony analysis, the use of GpIIb/IIIa<+> cells highly enriched for CFU-MK from rat bone marrow cells and plastic non-adherent cells obtained in the previous steps from GpIIb/IIIa+ cells, induced in human TPO the formation of a significant number of megakaryocyte colonies. For example, 34.5 megakaryocyte colonies were formed from 1000 Gplm/nia<+ > cells and 28.5 megakaryocyte colonies were formed from 20,000 plastic non-adherent cells in the presence of 20% (v/v) culture supernatant from transfected COS 1 cells. In addition, among the plastic non-adherent cell fraction containing different hematopoietic stem cells different from CFU-MK, only megakaryocyte colonies were formed in the presence of human TPO, strongly suggesting that human TPO exhibits its selective effect on stem cells of the megakaryocyte lineage. When the GpIIb/nia<+> cells obtained from rat bone marrow were cultured for 3 to 5 days in the presence of 20% (v/v) culture supernatant from transfected COS 1 cells in a liquid culture, the increase in ploidy was clearly observed on day 5 of culture.

(Example 39)

Construction of vector for use in the expression of a fusion protein (hereafter referred to as "GST-TPOQ-174)") of glutathione-S-transferase (GST) and human TPO (amino acid residues 1 to 174) in E. coli

To enable expression of human TPO in E. coli, an artificial gene encoding human TPO containing an E. coli preference codon was prepared. The nucleotide sequence of this DNA contains an amino-terminal methionine codon (ATG) at the -1 position for use in translation initiation in E. coli.

Synthetic oligonucleotides 1 to 12:

was prepared by each of the synthetic oligonucleotides 2 to 11 being phosphorylated with T4 kinase (manufactured by Pharmacia) in a solution composition of 1 mM ATP, 10 mM Tris-acetate, 10 mM Mg-acetate and 50 mM K-acetate. To obtain these synthetic oligonucleotides into 6 double-stranded DNA fragments, these synthetic oligonucleotides into combinations of 1 and 2,3 and 4,4, 5 and 6, 7 and 8,9 and 10 and 11 and 12, each combination was subjected to annealing in a solution composed of 10 mM Tris/HCl (pH 7.5), 10 mM MgCl2 and 50 mM NaCl. Then, 3 double-stranded DNA fragments from 1 and 2,3 and 4,5 and 6 and the other 3 double-stranded DNA fragments from 7 and 8,9 and 10 and 11 and 12, respectively, were treated with T4 ligase (manufactured by Life Technology), and these two reaction solutions were again treated with T4 ligase in the same way. The DNA fragment obtained by the ligation reaction was cut with BamHI (manufactured by Boehringer-Mannheim) subjected to 2% agarose gel electrophoresis to recover a fragment of about 390 to 400 bp then purified using the Prep-A-Gene DNA purification kit and subcloned into pUC18 which had previously been cut with XbaI and BamHI (E. coli DH50 was used as host). From the clones thus obtained, the selected clones with the artificial gene encoding human TPO and containing the following E. coli preference codon were selected by nucleotide sequence analysis and named pUC18(XB)(1-123). The DNA sequence of the coding chain is shown in sequence listing (SEQ ID NO: 9).

Single-stranded cDNA was synthesized from 1 ng of human normal liver-derived poly(A)<+ >RNA using oligo dT primer and PCR was performed using 0.1 volume of the cDNA synthesis reaction solution as a template. PCR reaction (incubation at 96°C for 2 minutes, repetition of a total of 30 reaction cycles, each cycle consisting of incubation at 95°C for 1 minute, at 58°C for 1 minute and 72°C for 1 minute and final incubation at 72 °C for 7 minutes) was performed with a volume of 100 µl using hTPO-C and hTPO-Z(EcoRl) as primers. Human TPO cDNA fragment thus obtained was cut with BamHI and EcoRI and subjected to 2% agarose gel electrophoresis to recover a fragment of about 600 bp in size which was then purified using the Prep-A-Gene DNA Purification Kit and subcloned into pUC18 which had been cut beforehand with EcoRI and BamHI (E. coli DH56 was used as host). A clone encoding the correct human TPO nucleotide sequence was selected by nucleotide sequence analysis and named pUC18(BE)(124-332). The oligonucleotide sequence of the primers used in the PCR reaction was as follows:

hTPO-C: 5--GGA GGA GAC CAA GGC ACA GGA-3'

(SEQ ID NO: 95) (positions 329 to 349 pHTF1 clone); and hTPO-Z(EcoRI): 5'-CCG GAA TTC TTA CCC TTC CTG AGA CAG ATT- 3' (SEQ ID NO: 96) (prepared by

addition of EcoRI recognition sequence to an antisense primer corresponding to positions 1.143 to 1.163 in the pHTF1 clone).

Clone pUC18(BE)(124-332) was cut with BamHI and EcoRI and subjected to 2% agarose gel electrophoresis to recover a human TPO cDNA C-terminal flanking fragment of about 600 bp in size which was then purified using Prep-A -Gene DNA Purification Kit and subcloned into pUC18(XB)(1-123) which had been cut beforehand with EcoRI and BamHI (E. coli DH50 was used as host). A clone thus obtained was named pUC18(XE)(1-332).

If the presence of human TPO activity in the human TPO peptide fragment of the amino acids in position 1 to 163 had been confirmed with an example in which different C-terminal deletion constructs of human TPO cDNA were prepared and expressed to measure in vivo human TPO activity, a expression vector containing this peptide fragment constructed. In this case, expression of the DNA sequence encoding GST-TPO(1-174) was performed.

When a nucleotide sequence corresponding to position 681 to 686 (amino acid residues in position 173 and 174) of the pHTF1 clone was recognized by restriction enzyme SacL, two terminal codons were introduced using two synthetic oligonucleotides using the recognition site of this restriction enzyme. That is, two synthetic oligonucleotides SSE1 and SSE2 were annealed in a solution composed of 10 mM Tris/HCl (pH 7.5), 10 mM MgCl2 and 50 mM NaCl and using the DNA Ligation Kit (manufactured by Takara Shuzo), it was thus obtained the double-stranded DNA fragment introduced into pUC18(XE)(l-332) from which the human TPO cDNA C-terminal fragment of about 480 bp had been removed by cutting it with SacI and EcoRI, thereby obtaining a clone pUC19 (XS)(1-174) (E. coli DH50 was used as host). The sequence of the synthetic oligonucleotide used here is as follows:

Clone pUC18(XS)(1-174) obtained above was cut with XbaI and EcoRI and subjected to 2% agarose gel electrophoresis to recover a fragment of about 600 bp in size encoding amino acid residues 1 to 174 of human TPO and this fragment was then purified using the Prep-A-Gene DNA Purification Kit and subcloned into Bluescript IISK<+> (manufactured by Stratagene) which had been cut beforehand with XbaI and EcoRI (E. coli DH50 was used as host). A clone thus obtained was named pBL(XS)(1-174). Similarly, clone pBL(XS)(1-174) was cut with XbaI and HindJJI to recover a fragment of about 550 bp in size encoding amino acid residues 1 to 174 of human TPO and the fragment was then purified using Prep- A-Gene DNA Purification Kit and subcloned into pCFM536 (EP-A-136490) which had been previously cut with XbaI and HindUI (E. coli DHSO was used as host). A clone thus obtained was named pCFM536/hT(1-174).

The following experiments were performed to effect the expression of GST-TPO(1-174) using pGEX-2T (manufactured by Pharmacia) which is an expression vector for glutathione-S-transferase (GST) fusion protein. PCR reaction (incubation at 96°C for 2 minutes, repetition of a total of 22 reaction cycles, each cycle comprising incubation at 95°C for 1 minute, at 41°C for 1 minute and 72°C for 1 minute and final incubation at 72° C for 7 minutes) was performed using pCFM536/hT(1-174) as a template and two PCR primers GEZ1 and GEX3. The human TPO coding fragment thus obtained was cut with Nael and EcoRI and subjected to 2% agarose gel electrophoresis to recover a fragment of about 550 bp in size, the fragment was then purified using the Prep-A-Gene DNA Purification Kit and cloned into in pGEX-2T which had been cut beforehand with EcoRI and SmaI (E. coli DH5 was used as host) and then a clone, named pGEX-2T/hT(1-174), encoding the correct human TPO nucleotide sequence, was selected by nucleotide sequence analysis for the preparation of transformant for use in the expression of GST-TPO(1-174). The oligonucleotide sequence of the primers used in the PCR reaction is as follows: GEX1:5'-ATC GCC GGC TCC GCC AGC TTG TGA C-3' (SEQ ID NO: 101) (prepared by adding the Nael recognition sequence of the sequence in position 21 to 39 in SEQ ID NO: 10); and

GEX3: 5'-GCC GAA TTC TCA TTA GAG CTC GTT CAG TGT-3'

(SEQ ID NO: 102) (an antisense primer corresponding to the sequence at positions 523 to 549 of SEQ ID NO: 10).

This expression plasmid contains a DNA sequence encoding the GST protein followed by a thrombin recognition peptide and a human TPO (amino acid residues 1 to 174). The DNA sequence encoding the thrombin recognition peptide and human TPO (amino acid residues 1 to 174) is shown in the sequence list (SEQ ID NO: 10).

(Example 40)

Expression of GST-TPOfl-174') in E. coli

The transformant prepared in Example 39 was grown overnight at 37°C in a shaker using 60 ml of LB medium containing 50 µg/ml ampicillin and a 25 ml portion of the resulting culture broth was added to 1000 ml of LB medium containing 50 µg/ml ampicillin and cultured at 37°C on the shaker until the OD at A( pQ reached 0.7 to 0.8. Then IPTG was added to the culture broth to the final concentration of 0.1 mM and the culture shaking was continued for an additional 3 h to induce the expression of GST-TPO( l-174).

(Example 41)

Purification of GST-TPO(l-174) expressed in E. coli and confirmation of its biological activity A 5.9 g aliquot of frozen cells of recombinant strains producing GST-TPO(l-174) derived from human TPO nucleotide sequence-encoding clone pGEX -2T/hT(1-174) was suspended in 10 ml of water and destroyed using a high pressure disintegrator. By subjecting the suspension to centrifugation, GST-TPO(1-174) was recovered in the precipitated fraction and most of the contaminating proteins, cellular components and the like were removed. Next, the thus prepared precipitated fraction containing GST-TPO(1-174) was suspended in 5 ml of water to which was then added, with stirring, 6 ml of 1 M tris buffer (pH 8.5), 120 ml of 10 M urea and 16 ml of water. After stirring for 5 minutes to dissolve the contents, the resulting solution was equally divided into four portions to carry out the following steps (1) to (4).

(1) One of the divided samples was diluted by a factor of 10 with 20 mM Tris buffer with a pH value of 8.5. To this was added reduced glutathione and oxidized glutathione to respective final concentrations of 5 mM and 0.5 mM, followed by overnight incubation at 4°C. The thus prepared mixture was subjected to centrifugation to recover GST-TPO(1-174) as the supernatant fluid which was then diluted by a factor of two with 20 mM sodium citrate buffer having a pH value of 5.5 and then adjusted to pH 5.5 with acetic acid. GST-TPO(1-174) in solution was loaded onto SP Sepharose Fast Flow (manufactured by Pharmacia Biotech, catalog no. 17-0729-01) cation exchange column that had been equilibrated beforehand with 20 mM sodium citrate buffer, pH 5.5. After washing the column with 20 mM sodium citrate buffer, pH 5.5, elution of GST-TPO(1-174) was performed using 20 mM sodium citrate buffer, pH 5.5, containing 50 mM sodium chloride. A 129 ml portion of the eluate was mixed with 2.6 ml of 1 M Tris buffer, pH 8.5, and the resulting mixture was pH 8.1 was applied to a glutathione sepharose 4B (manufactured by Pharmacia Biotech, catalog no. 17-0756-01) column to adsorb GST-TPO(1-174). After washing the column with PBS, elution of GST-TPO(1-174) was performed using 20 mM Tris buffer, pH 8.5, containing 10 mM of reduced glutathione. Resulting eluate was mixed with 37 NTH units of thrombin, set for 4 hours at room temperature, diluted with PBS by a factor of 10 and applied to a glutathione-pharose 4B column to adsorb cleaved GST and recover TPO(l-174) in the non- adsorbed fraction. The thus recovered unadsorbed fraction was diluted by a factor of 3 with 20 mM sodium citrate buffer, pH 5.5, applied to SP Sepharose fast Flow cation exchange column which had been equilibrated beforehand with the same buffer and the compound of interest was eluted with a linear density gradient of 0 to 500 mM sodium chloride dissolved in the same buffer.

(2) All operations in step (1) were performed in the presence of 0.1% polysorbate 80.

(3) One of the split samples was diluted by a factor of 10 with 20 mM Tris buffer with a pH of 8.5. To this was added reduced type glutathione oxidized type glutathione to the respective final concentration of 5 mM and 0.5 mM followed by standing overnight at 4°C. The thus prepared mixture was subjected to centrifugation to recover GST-TPO(1-174) as supernatant fluid which was then diluted by a factor of 2 with 20 mM sodium citrate buffer with a pH value of 5.5 and juiced to pH 5.5 with

acetic acid. GST-TPO(1-174) in solution was loaded onto SP Sepharose Fast Flow cation exchange resin that had been previously equilibrated with 20 mM sodium citrate buffer, pH 5.5. After washing the resin with 20 mM sodium citrate buffer, pH 5.5, elution of GSTTPO(1-174) was performed using 20 mM sodium citrate buffer, pH 5.5, containing 500 mM sodium chloride. The eluate was adjusted to pH 8 with 1 M tris buffer, pH 8.5, mixed with 320 NTH units of thrombin to remove GST polypeptide sequence via cleavage of thrombin recognition peptide linkers, left for 4 hours at room temperature, diluted by a factor of 5 with 20 mM sodium citrate buffer, pH 5.5, and then put on an SP Sepharose Fast Flow cation exchange column that had been equilibrated beforehand with the same buffer and the relevant compound was then eluted with a linear density gradient of 0 to 500 mM sodium chloride buffer dissolved in the same buffer.

(4) All operations in step (3) were performed in the presence of 0.1% polysorbate 80.

Each of the fractions eluted at sodium chloride densities of about 200 to 400 mM by SP Sepharose Fast Flow cation exchange column chromatography was thoroughly dialyzed with IMDM culture medium and then evaluated by the rat CFU-MK analysis system. The fractions were then analyzed by SDS-PAGE in the presence of a reducing agent, a band corresponding to a molecular weight of 19 kd was detected (purity 1 to 20%) as one of the major peptide bands of the fraction. When N-terminal sequence analysis of this band was performed by subjecting it to SDS-PAGE and PVDF membrane transfer according to the procedure described in Example 1, it was confirmed that this band contained a TPO sequence for expression as that of pGEX-2T/hT(l -174)-derived fusion protein. The deduced amino acid sequence of the resulting TPO was [GlyljTPO given the sequence of the thrombin recognition peptide linker and the known activity of thrombin on the linker.

(Example 42)

Construction of a vector for use in E. coli expression of human TPO (amino acid residues 1 to 163) with substitution in position 1 (Ser -> Ala) and position 3 (Ala

-> Choice)

Clone pUC18(XE)(1-332) prepared in Example 39 was cut with XbaI and EcoRI and subjected to 2% agarose gel electrophoresis to recover a fragment of about 1000 bp in size encoding amino acid residues 1 to 332 of human TPO and the fragment was then purified using the Prep-A-Gene DNA Purification Kit and subcloned into pBluescript IJSK+ (manufactured by Stratagene) which had been cut beforehand with XbaI and EcoRI (E. coli DH5α was used as host). A clone thus obtained was named pBL(XE)(1-332). Then, a region of clone pBL(XE)(1-332) from its BamHI recognition sequence to position 163 amino acids (positions 366 to 489) was changed to the E. coli dominant codon. Synthetic oligonucleotides 13 to 20 shown below were prepared and each pair of synthetic oligonucleotides 13 and 14, 15 and 16 and 17 and 18 were

phosphorylated in the same tube containing a solution composed of 0.1 mM ATP, 10 mM Tris-acetate, 10 mM Mg-acetate and 50 mM K-acetate. After further addition of 1/10 volume of a solution composed of 100 mM Tris/HCl (pH 7.5), 100 mM MgCl2 and 500 mM NaCl, the resulting solution was heated in a boiling water bath for 3 min and then cooled to room temperature to obtain double-stranded DNA fragment. Then the thus obtained 3 pairs of double-stranded DNA fragment comprising 13 and 14,15 and 16 and 17 and 18 were ligated using the DNA Ligation Kit (manufactured by Takara Shuzo) and PCR was performed using the thus ligated

the products as template and the synthetic oligonucleotides 19 and 20 as primers. The PCR product thus produced was then cut with BamHI and HindUI and subjected to 2% agarose gel electrophoresis to recover a fragment of about 130 bp in length using the Prep-A-Gene DNA Purification Kit. This was subcloned into pBL(XE)(1-332) which had been cut beforehand with BamHI and Hindm (E. coli DH5 was used as host). A clone with the following nucleotide sequences was selected from the clones thus obtained by sequence analysis and named pBL(XH)(1-163):

For the purpose of increasing the expression amount of human TPO (positions 1 to 163) and expected hydrolysis of the N-terminus by protease, an expression vector constructed for use in expression of a mutated type of human TPO (positions 1 to 163) (hereinafter referred to as "h6T (l-163)") with substitutions at position 1 (Ser to Ala) and position 3 (Ala to Val) of human TPO (positions 1 to 163) [(Ala<1>, Val<3>]TPO(l- 163)) and at the same time, which codes for Lys at the -1 position and Met at the -2 position [(Met"<2>, Lys-<1>, Ala<1>, Val<3>]TPO(l- 163)).Four synthetic oligonucleotides shown below are prepared and synthetic oligonucleotides 2-9 and 3-3 were phosphorylated with T4 kinase (manufactured by Pharmacia) in a solution composed of 1 mM ATP, 10 mM Tris-acetate, 10 mM Mg- acetate and 50 mM K-acetate. To turn these synthetic oligonucleotides into double-stranded DNA fragments, each pair of single-stranded DNA fragments 1-9 and 2-9 and 3-3 and 4-3 was annealed in a solution composed of 10 mM tris /HCl (pH 7.5), 10 mM MgCl2 and 50 mM NaCl. Then, the thus obtained two pairs of double-stranded DNA fragments comprising 1-9 and 2-9 and 3-3 and 4-3 were treated with the DNA Ligation Kit (manufactured by Takara Shuzo). The DNA fragment thus obtained by the ligation reaction was subcloned into pBL(XH)(1-163) which had been cut beforehand with XbaI and NruI and a clone correctly substituted with the following synthetic oligonucleotide was selected by a sequence analysis (E. coli DH50 was used as host) and given the name pBL(XH)h6T(1-163):

Clone pBL(XH)(1-163) was digested with XbaI and HindlH to recover a fragment of about 500 bp in size encoding amino acid residues 1 to 163 of mutant type human TPO and the fragment was then purified using Prep-A-Gene DNA Purification Kit and cloned into pCFM536 (EP-A-I36490) which had been cut beforehand with XbaI and HindUI (E. coli JM109 which had been transformed beforehand with pMW1 (ATCC No. 39933) which was used as host). A clone thus obtained was named pCFM536/h6T(l-163) and an E. coli strain with this expression vector was used as a transformant for use in the expression of mutant-type human TPO, namely h6T(l-163). This expression plasmid contains the DNA sequence shown in the sequence listing (SEQ ID NO: 11).

(Example 43)

Expression of h6T(l-163) in E. coli

Expression of the expression plasmid pCFM536 is controlled by the 'PL promoter which is itself under the control of the cl857 repressor gene. The transformants obtained in Example 42 were cultured overnight at 30°C in a residual application of 60 ml LB medium containing 50 mg/ml ampicillin and 12.5 µg/ml tetracycline and a 25 µl portion of the resulting culture broth was added to 1000 ml LB medium containing 5 ng/ml ampicillin and grown at 37°C on the shaker until the OD at AgQO reached 1.0 to 1.2. Then, about 330 ml of LB medium heated to 65°C was added to the culture broth to adjust the final medium temperature to 42°C and the shaking of the culture was continued for 3 hours at 42°C to induce the expression of h6T(l-163).

(Example 44)

Purification of h6T(l-163) expressed in E. coli and confirmation of its biological activity

A 3.6 aliquot of frozen cells of the h6T(1-163) producing recombinant strain was suspended in 10 ml of water and disrupted using a high pressure disintegrator. By centrifuging the suspension, the precipitated fraction was recovered and most of the contaminated proteins, cell components and the like were removed. The thus recovered precipitated fraction containing h6T(1-163) was suspended in 7 ml of water and 3 ml of 1 M Tris buffer (pH 8.5) was added to the suspension with stirring, followed by further addition of urea (final concentration, 8M), guanidine hydrochloride (final concentration 6M) and sodium N-lauroyl sarcosinate (final concentration 2%). After 5 to 20 minutes of stirring at room temperature to solubilize the contents, the resulting solution was diluted 10-fold with 20 mM Tris buffer with a pH value of 8.5 and subjected to protein refolding by overnight incubation at 4°C. In this case, glutathione and copper sulphate were used as additives in addition to air oxidation. Upon centrifugation, h6T-(1-163) was recovered in the supernatant fluid. When each of the recovered fractions was analyzed by SDS-PAGE in the presence of reducing agent, a band corresponding to a molecular weight of 18 kd was detected (purity, 30 to 40%) as the major protein band in each fraction. When a terminal sequence analysis of this band was performed by subjecting it to SDS-PAGE and PVDF membrane transfer according to the procedure described in Example 1, it was confirmed that this band contained a TPO sequence and was expressed as pCMF536/h6T(l-163)- derived mutation type TPO. When each fraction was thoroughly dialyzed against the IMDM culture medium and evaluated with the rat CFU-MK analysis system, TPO activity was found in all fractions in a dose-dependent manner.

EXAMPLE 45

Preparation of anti-TPO peptide antibody and detection of TPO by SDS-PAGE and western analysis

Successful preparation of anti-TPO peptide antibody will make it possible to detect TPO protein by immunological methods. Three regions that appear to be relatively useful as antigens (see below) were selected from the confirmed TPO amino acid sequences, and a quadruple-stranded multiple antigen peptide (MAP) type peptide was synthesized according to the procedure of Tam (Proe. NatLAcad. Sei. USA , 85, 5409-5413, 1988) using each of the selected regions, and then each of two rabbits was immunized 8 times with 100 µg each of the thus synthesized peptides to obtain respective antiserum samples.

Amino acid sequences included in the synthesized peptide antigens (a) Rat TPO (9-28) (peptide RT1 region)

(R in position 24 is originally S in the amino acid residues determined from the gene),

(b) Rat TPO (46-66) (peptide RT2 region)

(c) Rat TPO (163-180) (peptide RT4 region)

(LLD in positions 178-180 is originally TSG in the amino acid residues determined from the gene). Then these antiserum samples, anti-RT1 peptide and anti-RT2 peptide were first subjected to a protein A (PROSEP-A; manufactured by Bioprocessing Ltd, catalog no. 8427) column chromatography to yield IgG fractions (2.344 mg and 1.920 mg, respectively), and 54 mg and 32 mg portions of the respective fractions were biotinylated by their coupling with active type biotin (NHS-LC-Biotin II; manufactured by PIERCE, catalog no. 21336). A recombinant TPO containing sample was subjected to SDS-PAGE and then to electroblotting on PVDF or nitrocellulose membrane in the same manner as described in Example 1, and western analysis was performed in the usual manner using these biotinylated antibodies as primary antibodies.

The membrane after blotting was washed with a solution consisting of 20 mM Tris-HCl and 0.5 M NaCl (pH 7.5) (TBS) for 5 min and twice with 0.1% Tween 20-containing TBS (TTBS) in 5 minutes each and then treated with a blocking agent (BlockAce; manufactured by Dainippon Pharmaceutical, catalog no uk-B25) for 60 minutes. Next, the membrane was treated for 60 minutes with TTBS solution containing 10 Mg/ml biotinylated anti-TPO peptide antibody, 0.05% BSA and 10% BlockAce and then washed with TTBS for 5 minutes each. Then the membrane was treated for 30 min with a solution prepared by diluting alkaline phosphatase-labeled avidin (manufactured by Leinco Technologies, catalog no. Al 08) by a factor of 5000 with TTBS solution containing 10% BlockAce and washed twice with TTBS for 5 min and with TBS for 5 minutes and then color development was achieved using an alkaline phosphatase substrate (manufactured by Bio-Rad, catalog no. 170-6432). The Western analysis described above was performed at room temperature.

Based on the results, it was possible to recognize not only different recombinant rat TPO proteins expressed in COS 1 cells, but also different recombinant human TPO proteins expressed in the COS 1 cells and E.coli cells (plasmid pHTPl - or plasmid pHTFl-derived human TPO expressed in COS 1 cells and N-glycosidic bond cleaved product thereof, GST-TPO(l-174) expressed in E.coli and thrombin-cleaved product thereof and h6T(l-163) which is a mutation type of TPO expressed in E.coli, which has been described in the examples). In addition, the results have made it possible to analyze these different types of recombinant human TPO.

In addition to the above, the possibility of preparing avanti-human TPO peptide antibodies was confirmed. The antibodies obtained according to these techniques can be used not only for western analyses, but also for the purification of TPO by antibody columns and other commonly used antibody-using immunological methods.

Six regions of the human TPO amino acid sequence shown in SEQ ID NO: 7, which were expected to have relatively suitable antigenicity, were selected (see Table 4). Tetrameric multiple antigen peptide (MAP) type peptides, corresponding to the regions, were synthesized. Each peptide was immunized in two rabbits, eight times at a rate of 100mg/hour.

Apart from this, a monomeric peptide in which a cysteine residue had been attached to the C-terminus of each peptide region shown in Table 4 was also synthesized. This peptide was then used as a test antigen to determine an antibody titer using enzyme immunoassay. As a result of this, the increase of antibody titers was confirmed for the above generated sera so that these sera were used as antisera.

The antibodies were generated by immunizing each antigenic peptide in two rabbits to produce antisera directed against the peptide. The resulting two anti-HT1 peptide antibodies are referred to as anti-HT1-1 peptide antibody and anti-HT1-2 peptide antibody depending on the individual rabbits.

An example of purification of anti-HT1-1 antibodies will be illustrated below.

30 mg of a monomeric peptide of HT1 to which a cysteine residue had been attached was coupled to 12 ml of Sulfo-Link coupling gel (Pierce, Catalog No. 44895). A peptide solution containing the antigen was coupled for 15 minutes to the gel equilibrated with a coupling buffer (50 mM Tris, 5 mM EDTA-Na pH 8.5) with 6 volumes relative to the gel volume. After incubation for 30 minutes, the gel was then washed with coupling buffers with 3 volumes relative to the gel volume. The coupling buffer containing 0.05 M L-cysteine-HCl was added to the gel in an amount of 1 ml/ml gel to block unreacted radicals

over 15 minutes. After incubation for 30 minutes, the gel was washed with the coupling buffer at 8 volumes relative to the gel volume. All coupling reactions were carried out at room temperature. Consequently, the peptide was covalently bound to the gel at a coupling efficiency of 28.3%, and then an antigen column of 0.8 mg peptide/ml gel was formed. 76.7 ml (protein content 3620 mg) of the antiserum containing anti-HT1-1 peptide antibody that had been bled from one of the rabbits with a total amount of 78.4 ml was applied to the antigen column pre-equilibrated with 50 mM phosphate buffer (pH 8 ) containing 150 mM NaCl and 0.05% sodium azide, and then washed with the same buffer to<2>provide 105.9 ml of a flow-through fraction (protein content 3680 mg). Then, the adsorbed fraction was eluted with 0.1 M citrate buffer (pH 3.0) and immediately neutralized with 21.1 ml of 0.1 M carbonate buffer (pH 9.9), followed by concentration by ultrafiltration (using Amicon YM 30 membrane) to pass purified anti-HT1-1 peptide antibody in a solution of 50 mM phosphate buffer (pH 8) containing 150 mM NaCl and 0.()5% NaCl3. Likewise, the following antibodies were raised: anti HT1-2 peptide antibody (60.0 mg); anti-HT2-1 peptide antibody (18.8 mg); anti-HT2-2 peptide antibody (8.2 mg); and such. Anti-HT3 to-HT6 peptide antibodies could also be generated in a similar manner. 3 mg of affinity-purified anti-HT1-1 peptide antibody, anti-HT1-2 peptide antibody, anti-HT2-1 peptide antibody or anti HT2-2 peptide antibody was biotinylated by coupling to an activated biotin (Pierce, NHS-LC-Biotin II, Catalog No. 21336).

All of the above-mentioned purified antibodies were found to recognize and detect human TPO by western blotting (on nitrocellulose filter or PVDF) or SDS-PAGE of a recombinant human TPO standard partially purified from the culture supernatant of CHO cells where a gene coding for the amino acid sequence shown in Table 4 had been introduced and expressed.

Table 4

Human TPA amino acid sequences included in tetrameric MAP type synthetic peptide antigen

Human TPO (amino acid numbers 8-28) peptide HT1 region:

Human TPO (amino acid numbers 47-62) peptide HT2 region:

Human TPO (armonic acid numbers 108-126) peptide HT3 region:

Human TPO (amino acid numbers 172-190) peptide HT4 region:

Human TPO (amino acid numbers 262-284) peptide HT5 region:

Human TPO (amino acid numbers 306-332) peptide HT6 region:

EXAMPLE 46

Preparation of anti-TPO peptide antibody column

Because the anti-rat TPO peptide antibody obtained in Example 45 could recognize rat and human TOP molecules, immunoglobulin (IgG) fractions of anti-RT1 peptide antibody and anti-RT2 peptide antibody were coupled to a chemically activated gel support in the following manner to form anti- TPO peptide antibody columns. Each of the two antibodies used as materials was separately prepared from the sera of the two rabbits. The anti-RT1 peptide antibody preparation from two rabbits was designated anti-RT1-1 peptide antibody and anti-RT1-2 peptide antibody, respectively, and the anti-RT2 peptide antibody prepared from the other two rabbits was designated anti-RT2-1 peptide antibody and anti-RT2-2 peptide antibody, respectively . Because these antibodies were used separately for antibody column preparation, a total of 5 columns were obtained, i.e. two anti-RT1 peptide antibody columns (referred to as "anti-RT1-1 antibody column" and "anti-RT1-2 antibody column" below). Two anti-RT2 peptide antibody columns (hereinafter referred to as "anti-RT2-1 antibody column" and "anti-RT2-2 antibody column" and one antibody column (anti-RT1-2+ 2-1 mixture antibody column) which were formed by mixing avanti- RT1-2 peptide antibody with anti-RT2-1 peptide antibody and coupling the mixture with a gel.

Each antibody was dissolved in a solution of 50 mM Na phosphate and 0.15 M NaCl (pH 8.0) to a final concentration of 5 mg/ml (or 2.5 mg/ml each of mixed anti-RT1-2 peptide antibody and anti-RT2-1 peptide antibody in the case of anti-RT1 + 2 mixed antibody columns), and a 2.3 ml portion of the resulting solution was mixed with 1.54 ml by volume of a swollen formyl-activated gel (Formyl-Cellulofine, prepared by Chisso) and subjected to 2 h coupling reaction at 4 °C. To this was added 1.1 ml of a 10 mg/ml solution of a reducing agent (trimethylamine borane (TMAB); manufactured by Seikagaku Kogyo, catalog No. 680246) followed by 6 hours of further coupling reaction and subsequent centrifugation to isolate the gel part. A 10 ml portion of purified water was added to the gel portion which was re-isolated by centrifugation, and unreacted antibody molecules were removed by repeating this step four times. Then, the resulting gel was mixed with 4.6 ml of a blocking solution (0.2 M Na phosphate and 1 M ethanolamine, pH7.0) and 1.1 ml of the reducing agent solution and treated at 4 °C for at least 2 h to perform blocking of active groups of unreacted gel. Then the gel was washed with purified water and DPBS using a centrifuge, packed in a small column tube, washed with a 3 M potassium thiocyanate top solution and 1.0 M glycine-HCl (pH 2.5) solution, equilibrated with DPBS and then saved. 15 anti-TPO peptide antibody gels of the anti-RT1-1 antibody column, the anti-RT1-2 antibody column, the anti-RT2-1 antibody column, the anti-RT2-2 antibody column and the anti-RT1-2 + 2-1 mixed antibody column were the binding efficiency of the respective IgG fractions reached. 97.4%, 95.4%, 98.4%, 98.3% and 99.4%. In addition, amounts of IgG fractions linked per gel volume were 5.6 mg/ml gel, 5.8 mg/ml gel, 5.7 mg/ml gel, 5.7 mg/ml gel and 2.9 mg/ml gel. Since it was confirmed that the coupling efficiency and coupling amount do not significantly vary depending on the origin of the antibody and the difference in the antigens, the experiments shown in the following Example 47 were performed using these antibody columns.

EXAMPLE 47

Purification of TPO derived from culture supernatant obtained by transfection of COS 1 cells with expression vector pHTPl, using anti-TPO antibody column and then by reverse phase column chromatography and confirmation of its biological activity The M-07e assay system was used as an in vitro assay for the assessment of the following purification samples. Partially purified samples of TPO derived from the culture supernatant obtained in Example 35 from transfected COS 1 cells with expression vector pHTPl (fractions different from the main TPO fraction, i.e. a passed through fraction F1 and a fraction F3 eluted after F2 by washing with 20 mM Na citrate, pH 5.8 Macro-Prep Methyl HIC column and fractions Fl and F3 of the SP Sepharose Fast Flow column) were combined to form a TPO "subpool" fraction (5,463.79 ml; protein concentration 0.490 mg/ml; total protein 2,676 mg; relative activity , 12,000 and total activity, 32,380,000) and concentrated using an ultrafiltration unit (Omega Ultraset, nominal molecular weight cut-off of 8000, manufactured by Filtron), the solvent of the concentrated fraction was replaced with DPBS and then the sample with a small volume was treated with a YM 10 membrane equipped ultrafiltration unit (manufactured by Amicon) to make it into 120.2 ml of a DPBS solution containing 0.05% sodium azide. Then, all five anti-TPO peptide antibody gels formed in Example 46 were mixed together and formed into a single antibody column (1.6 cm in diameter and 4.8 cm in layer height, referred to as the anti-RTl + 2 mixture antibody column), and the TPO "subpool " fraction was applied to the column at a flow rate of 0.033 ml/min. After the end of sample loading, the elution was carried out with DPBS to collect eluates until the UV absorbance became sufficiently small, and the thus collected eluates were concentrated using a YM-10 membrane-equipped ultrafiltration unit (manufactured by Amicon) to obtain a continuous fraction Fl (82.62 ml; protein concentration, 14.3 mg/ml; total protein, I, 184 mg; relative activity, 67,500; and total activity, 79,900,000). Then, a column adsorbed fraction F2 (92.97 mL; protein concentration, 0.12 mg/mL; total protein, II, 1 mg; relative activity, 257,100; and total activity, 2,860,000) was eluted with an acidic elution solution (0.1 M glycine-HCl, pH 2.5). Despite the fact that flow-through fraction F1 still contained a TPO activity, the antibody column adsorbed TPO was further purified because it showed an approximately 20-fold increase in relative activity. Using 0.1% TFA as a developing solvent A and 1-propanol containing 0.05%

TFA as a developing solvent B, Capcell Pak Cl 300A (manufactured by

in Shiseido, catalog No. Cl TYPE:SG300A; 4.6 mm 8 diameter and 150 mm in layer height,

connected to a precolumn 4.6 mm in diameter and 75 mm in layer height) column chromatography was performed by mixing fraction F2 obtained from the antibody column with 1/10 volume of developing solvent B and applying the mixture at a flow rate of 0.4 ml /min on Capcell Pak Cl 300A column that has been pre-equilibrated with 20% B. After completion of sample loading, elution was performed for 5 min with 20% B and then for 50 min with a linear density gradient from 20% B to 40% B , and the eluates were collected in polypropylene tubes in 1 ml (2.5 min) portions.

A 2 µl portion (1/500 fraction) of the eluate in each tube was mixed with HSA and concen-

in trit by ultrafiltration to top up 0.25 ml of IMDM assay culture solution containing 0.02% HSA for use in the identification of TPO active fractions according to the assay. As a result, strong TPO activity was found in samples in tubes with numbers 20 to 23 (within the range of 28.5 and 32.5% as propanol concentration). In addition, when these samples were subjected to western analyzes according to the procedure described in Example 45, the presence of TPO was confirmed with an apparent molecular weight of 60,000 to 70,000 measured under a reduced condition based on TPC Ul molecular weight markers, and also the presence of other molecules with respective molecular weights of 32,000 to 43,000 and 20,000 to 30,000.

EXAMPLE 48

Confirmation of biological activity of TPA derived from culture supernatant obtained by transfection of COS 1 cells with expression vector pHTP1 and purified up to the step Capcell Pak C1300A column

TPO active fractions purified in Example 36, i.e. column numbers 20 to 23 (within the range of 28.5 and 32.5% as propanol concentration) of Capcell Pak Cl 300A column, were pooled, mixed with 0.21 ml of glycerol and then concentrated by centrifugation evaporation . This was mixed with 0.21 ml of 6 M guanidine hydrochloride solution, diluted to a volume of 1 ml with DPBS, buffer change with DPBS solution containing 0.01% HSA by Sephadex G25 column (NAP-10, manufactured by Pharmacia Biotech, catalog no. 17 -0854-01) and then mixed with 1.1 ml of DPBS solution containing 0.01% HSA, thereby finally obtaining a TPO active fraction FA (2.6 ml). In order to investigate the biological TPO activity of this sample, in vivo analysis was performed. The active fraction was subcutaneously administered to male ICR mice (8 weeks old), each group including 4 animals, daily for 5 days at a dose of 100 fal (with a total activity of 110,000 measured by the M-07e assay system) per animal. As a control, 100 μl of DPBS containing 0.01% HSA was subcutaneously administered in the same manner. Just before administration and on the next day of final administration, blood was collected from the eye socket to measure platelet counts using a microcell counter (F800, manufactured by Toa Iyo Denshi). In the TPO treated group, the platelet count increased by a factor of approximately 1.42 on average after completion of administration compared to the value before administration and was 1.23 times higher than was the case for the control group with a significant difference (p < 0.05 Student t-test ) between them. On the basis of these results, it was confirmed that the above human TPO produced by the mammalian cells has the ability to increase the platelet count in vivo.

EXAMPLE 49

Confirmation of biological activity of crude TPO fraction derived from 33 liters of culture supernatant obtained from transfected COS 1 cells with expression vector pHTP1 and partially purified by cation exchange column

(1) The M-07e analysis system was used as an in vitro analysis for the assessment of the following purification samples. In the same way as described in example 36, a total of 33 liters of serum-free culture supernatant was obtained by transfecting COS 1 cells with expression vector pHTP1 and filtered through a 0.22 µm filter to isolate the supernatant. This was concentrated by a factor of about 10 with an ultrafiltration unit (PLGC Pellicon Cassette, nominal molecular weight cut-off of 10,000, manufactured by Millipore) and mixed with about 1 mM in final concentration of a protease inhibitor p-APMSF to obtain a sample of 2.018 ml in volume (protein concentration, 3.22 mg/ml; total protein, 6.502 mg; relative activity, 66,000; total activity, 429,100,000). Then it became this

concentrated repeatedly using an ultrafiltration unit (Omega Ultraset, nominal molecular weight cut-off of 30,000; manufactured by Filtron), thereby obtaining a fraction of 30,000 or more in molecular weight (volume, 1,190 ml; protein concentration, 2.54 mg/ml; total protein , 3.020 mg; relative activity, 82,500; total activity, 249,000,000) and a fraction of 30,000 or less in molecular weight (volume, 2,947 ml; protein concentration, 0.471 mg/ml; total protein, 1.402 mg; relative activity, 4,500; total activity, 6,310,000). Presence of TPO activity in fraction 30,000 or less in molecular weight indicates a possibility that the culture supernatant contains TPO whose molecular weight has been reduced during expression in or secretion from the animal cells. In contrast, the fraction of 30,000 or more in molecular weight was buffer exchanged with 20 mm sodium citrate buffer (pH 6.1) and applied at a flow rate of 5 ml/min on SP Sepharose Fast Flow (2.6 cm in diameter and 29 cm in bed height, prepared by Pharmacia Biotech, catalog No 17-0729-01) column which had been equilibrated beforehand with 20 mm sodium citrate buffer (pH 6.1). After completion of sample application, the column was washed and eluted with 20 mm sodium citrate buffer (pH 6.1). Then, using 20 mm sodium citrate buffer (pH 6.1) as a developing solvent A and a solution consisting of developing solvent A and 1M NaCl as a developing solvent B, the elution was carried out at a flow rate of 3 ml/min with a linear gradient of from 0% B to 50% B for 215 min and then from 50% to 100% B for 20 min, and then the eluates were collected in polypropylene tubes in 30 ml (10 min) portions. When a portion of each of these eluates was examined in the M-07e analysis system to investigate the distribution of the activity, elution of the TPO activity was found within a wide range. As a consequence, fractions eluted with a NaCl concentration of 50 mM or lower, including the flow-through fraction, were pooled as a fraction F1, and the TPO fractions eluted with 50 to 1000 mM NaCl as a fraction F2. The volume of fraction F1 was 1.951 ml (protein concentration, 2.05 mg/ml; total protein 3,994 mg; relative activity, 13,500; total activity 53,900,000) and that of F2 was 649.8 ml (protein concentration, 1.11 mg/ ml; total protein, 721 mg; relative activity, 268,000; total activity, 193,000,000).

(2) Confirmation of biological activity of SP Sepharose Fast Flow fraction F2.

A 2.5 ml portion of SP Sepharose Fast Flow fraction F2 separated in the above step 1 was buffer exchanged with 3.5 ml DPBS solution containing 0.01% HSA by Sephadex G25 column (NAP-25, manufactured by Pharmacia Biotech, catalog No. 17 -0852-01), and biological TPO activity of this sample (protein concentration, 1.46 mg/ml) was examined in an in vivo assay. The active fraction was administered subcutaneously to male ICR mice (8 weeks old), each group containing 4 animals, daily for 5 consecutive days at a dose of 100 ni (with a total activity of 123,000 measured by the M-07e assay system) per animal. As a control, 100 µl of DPBS containing 0.01% HSA was administered subcutaneously in the same manner. Just before the administration and the day after the final administration, blood was collected from the eye socket ("eyeground") to measure the platelet count using a microcell counter (F800, manufactured by Toa Iyo Denshi). In the TPO administered group, the platelet count increased by a factor of approximately 1.29 on average after the end of administration compared to the value before administration and was 1.12 times higher than in the control group with a significant difference (p <0.0S, Student t-test) between them. These results demonstrated that the above human TPO produced by mammalian cells has the ability to increase platelet count.

EXAMPLE 50

Construction of recombinant vector, pDEF202-ghTPO for use in expression of human TPO chromosomal DNA in CHO cells

A vector pDEF202 was digested with the restriction enzymes KpnI and SpeI and then subjected to agarose gel electrophoresis to isolate a fragment containing the mouse DHFR minigene and SV40 polyadenylation content, and the expression vector pDEF202-ghTPO was obtained by ligation using T4 DNA ligase (prepared by Takara Shuzo) and the thus obtained fragment with a vector DNA which has been formed by removing an SV40 polyadenylation signal-containing region from a plasmid pEFHGTE by its treatment with the restriction enzymes KpnI and SpeI. This plasmid contains SV40 origin of replication, human elongation factor 1-a promoter, SV40 early polyadenylation region for transcription of human TPO gene, mouseD HFR minigene, pUC 18 origin of replication and α-lactamase gene (Amp<1>) and human TPO chromosomal DNA is linked to a downstream site of human elongation factor l-a promoter.

EXAMPLE 51

Expression of human TPO chromosomal DNA in CHO cells

CHO cells (a dhfr strain; Uralub and Chasin, Proe. Nati. Sei. USA, vol. 77, p.4216, 1980) were cultured by culturing them in an α-minimum essential medium (α-MEM(-), supplemented with thymidine and hypoxanthine) containing 10% fetal bovine serum using a 6 cm diameter dish (manufactured by Falcon), and the resulting cells were subjected to transfection using a calcium phosphate method (Cell Phect, manufactured by Pharmacia). 10 ng of the plasmid pDEF202-ghTPO prepared in Example 50 was mixed with 120 µl of buffer A and 120 µl of H 2 O and the resulting mixture was incubated at room temperature for 1 10 minutes. This solution was further mixed with 120 µL of buffer B and incubated at room temperature for a further 30 minutes. The DNA solution formed in this way was placed in the plate and cultivated for 6 hours in a CO2 incubator. After removing the medium and washing the resulting plate twice with α-MEM(-), α-MEM(-) containing 10% dimethyl sulfoxide was added to the dish and treated for 2 minutes at room temperature. Then, non-selection medium (α-MEM(-) op.cit. supplemented with hypoxanthine and thymidine) containing 10% dialyzed fetal bovine serum was added and cultured for 2 days, and then selection was performed using a selection medium (α-MEM (-), free of hypoxanthine and thymidine) containing 10% dialyzed fetal bovine serum. Selection was performed by treating the cells with trypsin, dispensing the treated cells in a 6 cm diameter dish into 5 10 cm diameter dishes or 20 24-well dishes and continuing to grow the cells while changing the selection medium every other day. Presence of human TPO activity was confirmed in the culture supernatant in dishes or wells where proliferation of cells was observed when human TPO activity was measured according to the Ba/F3 analysis. In this case, transfection of CHO cells can also be achieved by performing cotransfection of CHO cells with pEFHGTE and pMG1.

EXAMPLE 52

Construction of vector for use in expression in E.coli of human TPO protein (amino acid residues in positions 1 to 163) (referred to below as "hMKT(1-163)" where a Lys residue and Met residue were respectively added to -1 and - 2 the positions and confirmation of the expression

To achieve expression of a protein in which the 1- and 3-position amino acid residues are the same as those of human TPO protein (the amino acid residues in positions 1 to 163), a vector pCFM536/hMKT(1-163), also referred to as [Met" <2>, Lys"<1>] TPO(1-163), was constructed for use in the expression of hMKT(1-163) in E.coli in the same manner as described in Example 42 using the following synthetic oligonucleotides 1- 13.2-13.3-3 and 4-3:

This expression plasmid contains a DNA sequence shown in the sequence listing (SEQ DD NO: 12). Then expression of hMKT(1-163) was induced in the same way as described in Example 43.

When the expression protein thus obtained was subjected to SDS-PAGE, transferred onto a PCDF membrane and then subjected to N-terminal amino acid sequence analysis, it was confirmed that this protein contains the amino acid sequence of hMKT(1-163) to be expressed.

EXAMPLE 53

Refolding of human TPO, h6T(l-163), derived from human TPO expressed in E.coli using guanidine hydrochloride and glutathione, and purification and confirmation of its biological activity

A 1.2 g portion of frozen cells of the h6T(1-163) producer combinatorial strain prepared in Example 43 was suspended in 3 ml of water and disrupted using a high pressure disintegrator. By subjecting the suspension to centrifugation, the precipitated frac.? tion isolated and most of the contaminated proteins, cell components and the like were removed. The isolated precipitated fraction containing h6T(1-163) was suspended in 4 ml of final volume water. A 1 ml portion of 1 M Tris buffer (pH 8.5) was added to the suspension with stirring followed by further addition of 20 ml of 8 M guanidine hydrochloride and then 5 minutes of stirring at room temperature to dissolve the contents. The resulting solution was diluted 10 times with 20 mM Tris buffer (pH 8.5), 5 mM reduced-type glutathione and 0.5 mM oxidized-type glutathione were dissolved in the diluted solution by stirring, and then the solution thus formed was the solution incubated overnight at 4 °C. By centrifugation, h6T(1-163) was isolated in the supernatant fluid. A 160 ml portion of the supernatant fluid thus obtained was concentrated using YM 10 ultrafiltration membrane (manufactured by Amicon) and subjected to buffer exchange with DPBS to obtain 3.4 ml final volume of a fraction (protein concentration, 2.18 mg/ml ). When this fraction was thoroughly dialyzed against IMDM culture medium and assessed according to the rat CFU-MK analysis system, strong TPO activity (relative activity, 58,000) was detected.

The TPO active fraction formed above was subjected to an in vivo analysis. The active fraction was administered subcutaneously to IC male mice (7 weeks old), each group containing 4 animals, daily for 5 consecutive days at a dose of 170 fil (with a total activity of 22,000 measured according to rat CFU-MK assay systems) per animal. As a control, 170 µl of DPBS was administered subcutaneously to each of 6 animals in the control group in the same manner. Just before the administration and on the next day and 3 days after the end of the administration, blood was collected from the eye socket to measure the platelet count using a microcell counter F800, (manufactured by Toa Iyo Denshi). In the TPO-administered group, the platelet count increased by a factor of approximately 2.15 on average on the next day of the end of administration compared to the value before administration, and such an effect remained unchanged even after 3 days of completion of administration showing a 2.13-fold higher value. The platelet counts on the next day and 3 days after the end of administration were 1.73 and 1.80 times higher in the TPO-treated group than in the control group with a significant difference (p<0.001, Stutent t-test) between them. These results demonstrated that human TPO produced by E.coli and formed according to the above method has the ability to increase platelet counts in vivo.

EXAMPLE 54

Refolding of human TPO, h6T(l-163) expressed in E.coli, using sodium N-lauroyl sarcosinate and copper sulfate, and purification and confirmation of its biological activity

A 0.6 g portion of frozen cells of the h6T(1-163) producing recombinant strain produced in Example 43 was suspended in 3 ml of water, disrupted using a high pressure disintegrator and then subjected to centrifugation to isolate the precipitated fraction. After suspending the precipitated fraction in 3.1 mL of water, 0.19 mL of 1 M Tris buffer (pH 9.2), 11.25 µl of 1 MDTT, 38 µl of 0.5 M EDTA, and 0.38 mL of 10% sodium deoxy -cholate solution was added to the suspension with stirring, and the resulting mixture was stirred at room temperature for 40 minutes. This was then subjected to centrifugation to isolate the precipitated fraction and remove most of the contaminating proteins, cellular components and the like. The resulting precipitate was suspended in 40 ml of water and subjected to centrifugation to isolate the precipitated fraction containing h6T(1-163). The precipitated fraction thus isolated was suspended in 3.8 ml of water and 0.2 ml of 1 M Tris buffer (pH 8) and 1 ml of 10% sodium N-lauroyl sarcosinate was added to the suspension with stirring followed by 20 minutes of stirring at room temperature to solubilize the contents. The resulting solution was mixed with 5 n 1% copper sulfate and stirred overnight (about 20 hours at room temperature). The resulting solution was subjected to centrifugation to isolate the supernatant fraction containing h6T(1-163) and a 5 ml portion of the supernatant was mixed with 5 ml water and 10 ml 20 Flow Cation Exchange Column Chromatography was thoroughly dialyzed against IMDM culture medium and assessed according to rat CFU-MK the analysis system became strong TPO activity

(relative activity 19,000,000) found in a fraction eluted with a NaCl concentration of approximately 100 mM. This fraction was concentrated 1.6-fold (2.5 ml; protein concentration, 25 ng/ml) using an ultrafiltration unit (Ultra Free CL, nominal molecular weight cut-off of 5,000, manufactured by Millipore, catalog No. UFC4LCC25). When the thus concentrated fraction was analyzed by SDS-PAGE in the presence of a reducing agent, a band corresponding to TPO was detected as the main band (purity 70 to 80%).

The TPO active fraction formed according to the above procedure was examined in an in vivo assay. The active fraction was administered subcutaneously to ICR male mice (8 weeks old), each group including 4 animals, daily for 5 consecutive days in a dose of 100 µl (with a total activity of 47,500 measured according to the rat CFU-MK assay system) per animal. As a control, 100 µl of 20 mM Tris buffer (pH 7.7) containing 100 mM NaCl was administered subcutaneously in the same manner. Just before the administration and the next day after the administration was completed, blood samples were collected from the fundus to measure the platelet count using a microcell counter (F800, manufactured by Toa Iyo Denshi). ITPO-administered group increased platelet counts by a factor of approximately 1.73 on average after the end of administration compared to the pre-administration count and was 1.59 times higher than in the control group with a significant difference (p < 0.01, Student t-test) between them. These results demonstrated that human TPO produced by E.coli and formed according to the above procedure has the ability to increase platelet counts in vivo.

EXAMPLE 55

Large-scale cultivation of CHO cells

Large-scale cultivation of human TPO-producing CO cell line (CHO 28-30 cell, resistant to 25 nM MTX) which had been obtained by transfecting human TPO expression plasmid pDEF202-hTPO-P1 into CHO cells in Example 32 was carried out as follows. The CHO cell line was grown and proliferated in DMEM/F-12 culture medium (GD3CO) containing 25 nM MTX and 10% FCS. After the cells were detached with trypsin solution, 1 X 10<?> were inoculated into a roller bottle (ex Falcon, Falcon 3000) containing 200 ml of the same medium and then cultured at 37°C at a rotation speed of 1 rpm for 3 days. After 3 days of culture, the culture supernatant was removed by suction and adherent COH cells were then rinsed with 100 ml of PBS. 200 ml DMEM/F-12 medium (GD3CO) containing hvscale culture of human TPO-producing CO cell line (CHO28-30 cell, resistant to 25 nM MTX) which had been obtained by transfection of human TPO expression plasmid pDEF202-hTPO-Pl into CHO cells in Example 32 was carried out as follows. The CHO cell line was grown and proliferated in DMEM/F-12 culture medium (GJJ3CO) containing 25 nM MTX and 10% FCS. After the cells were detached with trypsin solution, 1X10^ was inoculated into a roller bottle (ex Falcon, Falcon 3000) containing 200 ml of the same medium through a 0.22 µm filter to obtain the filtrate which was then concentrated on an ultrafiltration unit (PLTK Pellicon cartridge, nominal molecular weight cleavage 30,000, ex Millipore). After replacing the solvent of the concentrate with pure water, p-APMSF (Wako Pure Chemicals) and Pefabloc SC (Merck) which are proteinase inhibitors were added to the resulting aqueous solution at final concentrations of 1 mM and 0.35 mM, respectively, to obtain 3628 ml of the fraction with a molecular weight of 30,000 or higher (protein concentration: 1.60 mg/ml; total protein: 5805 mg; relative activity; 1,230,000; total activity: 7,149,000,000). The fraction was then subjected to western blot analysis as described in example 45. As a result, the presence of TPO protein emerged at a molecular weight range between 66,000 and 100,000. Upon examination in more detail, it was discovered that TPO species with a lower molecular weight than the above-mentioned TPO were included in the flow-through fraction.

The ultrafiltrate containing TPO with a molecular weight of no more than 30,000 was separately concentrated using an ultrafiltration unit (PLGC Pelliconkassette nominal molecular weight separation 10,000, Millipore), to obtain a 1901 ml TPO fraction between molecular weight (protein concentration: 0.36 mg/ml; total protein: 684 mg; relative activity: 245,500; total activity: 167,900,000). This shows that TPO species with low molecular weight were formed during the cell culture.

Then, to 3614 ml of the fraction containing TPO with a molecular weight of 30,000 or more, 764 g of ammonium sulfate and 144.5 ml of 0.5 M sodium citrate buffer (pH 5.5) were added to form 4089 ml of the solution containing 1.41 M ( final concentration) ammonium sulfate and 17.7 mM (final concentration) sodium citrate buffer. After insoluble compounds were removed from the solution by centrifugation, the clear solution was obtained.

The clear solution was then applied to a Macro-Prep Methyl HIC column (Bio-Rad, Catalog No. 156-0080; diameter 5 cm, layer height 24.5 cm) which had previously been equilibrated with 20 mM sodium citrate buffer (pH 5.5) containing 1.2 M ammonium sulfate at a flow rate of 25 ml/min. After this was finished, the elution was carried out with 20 mM sodium citrate buffer (pH 5.5) containing 1.2 M sodium sulfate. The eluted fraction was then concentrated on an ultrafiltration unit (ex Filtron, Omega Ultrasette, nominal molecular weight cut-off of 8,000) to obtain a yeast recirculation fraction Fl (4455 ml; protein concentration: 0.400 mg/ml; total protein: 1780 mg; relative activity: 1,831,000) .

Then, 20 mM Na citrate (pH 6.0) as an elution buffer was passed through the column to obtain an eluate which was then concentrated using the same ultrafiltration unit (ie Omgea Ultrasette, nominal molecular weight cut-off 8,000) to collect a fraction F2 (1457 ml ; protein concentration: 0.969 mg/ml; total protein: 1411 mg; relative activity: 1,715,000). IF2, the presence of a protein with a molecular weight from 66,000 to 100,000 was confirmed on SDS-PAGE. As a result of Western analyses, it appeared that the protein was TPO. F2 eluted from the Macro-Prep Methyl HIC column (1443 ml) was applied to SP Sepharose Fast Flow (Pharmacia Biotech, Catalog No. 17-0829-01; diameter 5 cm, layer height 12 cm) which had previously been equilibrated with 20 mM sodium citrate buffer (pH 6.0) at a flow rate of 15 ml/min. After this was completed, the elution was carried out with 20 mM sodium citrate buffer (pH 6.0) containing 50 mM NaCl. The eluate was collected and the beads were concentrated using the same ultrafiltration unit (ie, Omgea Ultrasette, nominal molecular weight cut-off 8,000) to collect a fraction F2 (1457 ml; protein concentration: 0.969 mg/ml; total protein: 1411 mg; relative activity: 1,715,000). IF2, the presence of a protein with a molecular weight from 66,000 to 00,000 was confirmed on SDS-PAGE. As a result of Western analyses, it appeared that the protein was TPO.

F2 eluted from the Macro-Prep Methyl HIC column (1443 ml) was applied to SP Sepharose Fast Flow (Pharmacia Biotech, Catalog No. 17-0829-01; diameter 5 cm, layer height 12 cm) which had previously been equilibrated with 20 mM sodium citrate buffer (pH 6.0) at a flow rate of 15 ml/min. After this was completed, the elution was carried out with 20 mM sodium citrate buffer (pH 6.0) containing 50 mM NaCl. The eluate was collected and designated fraction F1 (3.007 ml; protein concentration: 0.226 mg/ml; total protein 679 mg; relative activity: 88,830). Then the elution buffer was replaced with 20 mM sodium citrate buffer (pH 5.4) containing 750 mM NaCl to collect eluted fraction F2 (931 ml; protein concentration: 0.763 mg/ml; total protein: 710 mg; relative activity: 5,558,000 ) which was then concentrated to 202 ml using ultrafiltration units (Filtron, Omega Ultrasette, nominal molecular weight cut-off 8,000; and Amicon, YM3 membrane).

Then the concentrated TPO activity-containing fraction F2 (197 ml) was applied at a molecular weight of 66,000 to 1,000,000; F2 a TPO molecular species with a molecular weight of 32,000 to 60,000; and F3 a TPO species with a molecular weight of 32,000 to 42,000. All the TPO molecules had TPO activity.

Based on the result of N-terminal amino acid sequencing of the TPO molecule with a molecular weight of 66,000 to 100,000, it was confirmed that the TPO molecule had the amino acid sequence of a human TPO gene-encoded protein.

In addition, the TPO molecule with a molecular weight of 66,000-100,000 was subjected to enzymatic cleavage experiments using glycosidase enzymes consisting of N-glycanase (ex Genzyme, Catalog No. 1472-00), neuraminidase (Nakarai-Tesqu, Catalog No. 242-29SP) , endo-1-N-acetylgalactosaminidase (Seikagaku Kogyo, Catalog No. 100453) and O-glycosidase (Boehringer Mannheim Biochemica, Catalog No. 1347101) individually or in combination, and then SDS-PAGE analyses. As a result, the polypeptide portion of TPO was found to have a molecular weight of approximately 36,000 as expected from the theoretical molecular weight and is a glycoprotein with both N- and O-linked sugar chains.

EXAMPLE 57

Purification of human TPO from human TPO-producing CHO cell line

(1) To 11 of serum-free culture supernatant of the CHO cells obtained in the same way as in example 55, 211.4 g of ammonium sulfate was added and then the mixture was filtered through a 0.2 nm filter (ex Gelman Science, Catalog No. 12992). The filtrate obtained was applied to a Macro-Prep Methyl HIC column (Bio-Rad, Catalog No. 156-0081, diameter 50 mm, layer height 90 mm) consisting of N-glycanase (ex Genzyme, Catalog No. 1472-00), neuraminidase (Nakarai-Tesqu, Catalog No. 242-29SP), endo-1-Nacetylgalactosaminidase (Seikagaku Kogyo, Catalog No. 100453) and O-glycosidase (Boehringer Mannheim Biochemica, Catalog No. 1347101) individually or in combination, and then SDS -PAGE analyses. As a result, the polypeptide portion of TPO was found to have a molecular weight of approximately 36,000 as expected from the theoretical molecular weight and is a glycoprotein with both N- and O-linked buffer (pH 6.4) containing 5% ethanol (referred to as "development solvent A ") for 15 min, the elution was performed with a 66-min linear gradient from Development Solvent A to 10 mM Tris buffer (pH 6.4) containing 94% ethanol (referred to as "Development Solvent B"). The chromatogram obtained is shown in Fig. 15. As a result of SDS-PAGE

the analysis revealed the molecule which has a molecular weight of 65,000-100,000 which was expected to be TPO and which corresponded to the fraction with retention time 68-72 min after addition of sample as a single band on the SDS-PAGE gel (see fig 16). Western analyzes further showed that the protein obtained is TPO.

N-terminal amino acid analysis of the protein sample showed that the protein had the amino acid sequence of the human TPO gene encoded protein.

EXAMPLE 58

Production of recombinant virus for human TPO expression in insect cells

Plasmid pHTP1 prepared in Example 30 was digested with the restriction enzymes EcoRI and Noti and then subjected to 1% agarose gel electrophoresis to obtain an approximately 1200 bp band which was then purified using the Prep-A-Gen DNA purification kit. Purified DNA was then ligated into a transfer vector pVL 1393 (Invitrogen) pretreated with the same restriction enzymes followed by transformation into competent tall E.coli DH5 (Toyo Boseki). From colonies obtained, a clone pVL1393/hTPO containing the full-length coding region of human TPO cDNA was selected. Plasmid DNA was then generated from the clone using the method described in Molecular Cloning (Sambrook et al, Cold Spring Harbor Laboratory Press, 1989) followed by transfection into insect cell Sf 21 (Invitrogen) using the BaculoGoldTM transfection kit (Farmingen), and transfected cells were grown in Sf-900 medium (Lifetechnology) at 27°C for 4 days to isolate the supernatant containing virus. The supernatant was then diluted to: 10/9 ~ 1:10/5 and approximately 7 X 10/5 Sf 21 cells were infected with 1 ml of the diluted supernatant at 27 °C for 1 hour in a 35-mm (in diameter) plate . After removal of the supernatant, 1% warm agarose in Sf-900 medium was placed in the culture and solidified, followed by 6-day cultivation at 27°C in a humidified atmosphere. When single plaques formed were picked up and the virus was released in 200 µl of Sf-900 medium. The Sf 21 cells were infected with the obtained viral clone in a 24-well dish to proliferate virus. A portion of the virus-containing fluid from single plaques was treated with phenol/chloroform followed by ethanol precipitation to isolate viral DNA which was then used as a template to perform PCR using primers specific for human TPO cDNA. Recombinant virus-containing human TPO cDNA was selected based on amplification of specific DNA fragment. Then, Sf 21 cells were infected with supernatant containing recombinant virus containing human TPO cDNA to proliferate virus.

EXAMPLE 59

Expression of human TPO in Sf 21 insect cells and identification of TPO activity

The Sf 21 cells were grown in a 175 cm culture flask to approximately 80% confluence followed by infection with human TPO cDNA containing recombinant virus produced in Example 58 at 27°C for 1 hour after which the resulting cells were grown in Sf-900 medium at 27 °C for 4 days to isolate the culture supernatant. The obtained culture supernatant was replaced with IMDM culture medium on the NAPTM-5 column (Pharmacia). Significant TPO activity in the results was detected in a dose-dependent manner in both rat CFU-MK analysis and M-07e analysis. Recombinant human TPO expressed in the Sf21 cells was identified using Western analyzes as described in Example 45.

EXAMPLE 60

Refolding of variant human TPO, h6T(l-163) derived from clone (pCFM536/h6T(l-163) containing a human TPO base sequence via expression thereof in E.coli using sodium N-lauroyl sarcosinate and cupric sulfate and purification of h6T( l-163)

30 g of frozen h6T(1-163)-producing recombinant microorganism formed in Example 43 was suspended in 300 ml of water, destroyed in a high pressure destructor (10,000 psi, Rannie High Pressure Laboratory) and then centrifuged to isolate a sedimented fraction. The settled fraction was suspended in 90 ml of water and water was added to the total amount of 150 ml with stirring. Then, 9 ml of 1 M Tris buffer (pH 9.2), 540 µl of 1 M DTT, 1.8 ml of 0.5 M EDTA and 18 ml of 10% sodium deoxycholic acid were added to the mixture with stirring followed by 30 min of stirring at room temperature. The sedimented fraction was isolated by centrifugation and most of the contaminating proteins and microorganism components were removed. To the sediment was added 180 ml of 5 mM DTT to obtain a suspension which was then centrifuged to isolate the sedimented fraction containing h6T(1-163). To the resulting sediment was added 300 ml of water to obtain a suspension to which water was added to a liquid amount of 570 ml with stirring. 30 ml of 1 M Tris buffer (pH 8), 150 ml of 10% sodium N-lauroylsarcosine was added to the mixture at room temperature with 20 min stirring to dissolve h6T(1-163). To this was further added 750 ul of 1% cupric sulphate and the mixture was stirred overnight (approximately 201) at room temperature. After the supernatant fraction containing h6T(1-163) was isolated by centrifugation, 750 ml of water and 1,500 ml of 20 mM Tris buffer (pH 7.7) were added to 750 ml of the supernatant followed by the addition of 3 ml of polysorbate 80 (Nikko Chemicals) with stirring. To the obtained solution was added 600 g of Dowex 1-X4 (20-50 mesh, chloride form) ion exchange resin followed by 90 min stirring at room temperature. Following isolation of the non-adsorbed fraction using a glass filter, the ion exchange resin was washed with 750 ml of 20 mM Tris buffer (pH 7.7). The combined solution of the non-adsorbed and washed fractions was adjusted to pH 9.2 with 2 N sodium hydroxide after which it was applied to a Q Sepharose Fast Flow anion exchange column (ID 5 x m X 10 cm) equilibrated with 20 mM Tris buffer (pH 9.2) containing 0.1% polysorbate 80 to isolate a non-adsorbed fraction. The pH of the obtained non-adsorbed fraction was adjusted to 7.2 with hydrochloric acid, applied to SP Sepharose Fast Flow cation exchange column (ID 5 cm X 10 cm) equilibrated with 20 mM Tris buffer (pH 7.2) containing 0.1% polysorbate 80 and then eluted with a linear gradient from 0 M to 500 mM NaCl in the same buffer. The eluted fractions were subjected to SDS-PAGE analysis to collect a TPO fraction to which trifluoroacetic acid was added to a concentration of 0.1%. After the obtained solution was applied to Capcell Pak 5 nm 300A Cl clone (ID 2.1 cm X 5 cm X 2, Shiseido) a reverse phase HPLC was performed by the linear gradient elution method by increasing the l-propanol concentration in 0.1% trifluoroacetic acid . The eluate was analyzed on SDS-PAGE in the absence of a reducing agent. As a result, three fractions were separated based on the elution position in reverse phase HPLC and each fraction was diluted three times with 0.1% trifluoroacetic acid. Each diluted fraction was applied to the above reverse phase HPLC column in a similar manner. The resulting three TPO fractions which were designated Fr. S-a, Fr. S-b and Fr. S-c depending on the elution order on reverse phase HPLC (see Fig. 17) was analyzed on SDS-PAGE in the absence of a reducing agent. As a result, a single band was detected with a molecular weight of approximately 18 kDa (Fr. S-a), approximately 19 kDa (Fr. S-b) or 18 kDa (Fr. S-c) (see Fig. 18). According to amino acid analyses, each of the amino acid compositions determined was almost in agreement with the corresponding theoretical value assessed from the sequence information. In addition, the results of N-terminal amino acid analysis showed that they were the expected sequences. The protein amount of each fraction was determined by amino acid-

analyzes were 0.64 mg (Fr. S-a), 1.81 mg (Fr. S-b) or 3.49 (Fr. S-c). After complete dialysis of each fraction against IMDM medium, M-07e analysis was performed. As 64 mg (Fr. S-a), 1.81 mg (Fr. S-b) or 3.49 (Fr. S-c). After complete dialysis of each fraction against IMDM medium, M-07e analysis was performed. As a result, the TPO relative activity of each fraction was approximately 1,620,000 (Fr. S-a), 23,500,000 (Fr. S-b) or 746,000,000 (Fr. S-c).

EXAMPLE 61

Refolding of variant human TPO, h6T(l-163) derived from clone pCFM536/h6T containing human TPO base sequence via expression in E.coli using guanidine hydrochloride and cysteine-cysteine, and purification of h6T(l-163) 50 g h6T( 1-163)-producing frozen recombinant microorganism prepared in Example 43 was added to 500 ml of water and suspended, disrupted using a high pressure apparatus (10,000 psi, Rannie High Pressure Laboratory), and then centrifuged to isolate a sediment fraction. The sediment fraction was suspended in water and the liquid was then adjusted to 250 ml by adding water with stirring. To this was then added 15 ml of 1 M Tris buffer (pH 9.2), 900 ni M DTT, 3 ml of 0.5 M EDTA and 30 ml of 10% sodium deoxycholic acid and stirred at room temperature for 30 min. After centrifugation, the sedimented fraction was isolated while most of the contaminating proteins, microorganism components and the like were removed. To the sediment was added 300 ml of 5 mM DTT and suspended and then the sedimented fraction containing h6T(1-163) was isolated by centrifugation. The isolated sedimented fraction containing h6T(1-163) was suspended in water in a total volume of 104 ml. To the suspension was added 20 ml of 1 M Tris buffer (pH8.5) and then 376 ml of 8 M guanidine hydrochloride with stirring, and then stirred at room temperature for 10 min to dissolve h6T(1-163). To this was added 2,500 ml of 20 mM Tris buffer (pH 8.5) containing 0.1% polysorbate 80 and then 2,000 ml of 20 mM Tris buffer (pH 8.5) containing 1 M guanidine hydrochloride with stirring, followed by the addition of 5 mM cysteine and 0.5 mM cystine. The solution thus obtained was incubated overnight at 4°C and then centrifuged to isolate h6T(1-163) in the supernatant which was then concentrated using PrepScale UF cartridge PLDC ultrafiltration membrane (Millipore). The buffer of the concentrate was replaced with 20 mM Tris buffer (pH 9.2) containing 0.1% polysorbate 80 until a final volume of 1,000 ml was obtained. The obtained solution was applied to Q Sepharose Fast Flow anion exchange column (ID 5 cm X 10 cm) equilibrated with 20 mM Tris buffer (pH 9.2) containing 0.1% polysorbate 80, and then eluted with a linear gradient from 0 M to 500 mM NaCl in the same buffer whereby the fraction (240 ml) eluted at about 20 mM - and about 150 mM NaCl was isolated. The resulting fraction was diluted four times with 20 mM Tris buffer (pH 7.2) to obtain a total volume of 960 ml, and then adjusted to pH 7.2 with acetic acid, applied to SP Sepharose Fast Flow cation exchange column (ID 5 cm X 10 cm) equilibrated with 20 mM Tris buffer (pH 7.2) containing 0.1% polysorbate 80, and then eluted with a linear gradient from 0 M to 500 mM NaCl in the same buffer. The eluate was analyzed on SDS-PAGE to collect a TPO fraction. Trifluoroacetic acid was added to the TPO fraction until the final volume was 0.1%. The resulting solution was applied to Capcell Pak 5 nm 300A Cl column (ID 2.1 cm X 5 cm X2, Shiseido) and reverse phase HPLC was then performed by the linear gradient elution method with increasing 1-propanol concentration in 0.1% trifluoroacetic acid. The eluate was analyzed by SDS-PAGE in the absence of a reducing agent and fractionated into two fractions depending on the eluted position in reverse phase HPLC. The two TPO factions were designated Fr. G-a and Fr. G-d based on their elution order in reverse phase HPLC (see Fig 17). Each fraction was then analyzed on SDS-PAGE in the absence of a reducing agent. As a result, a single band was detected at a molecular weight of about 18 kDa (Fr. G-a) or about 32 kDa (Fr. G-d) (see Fig. 18). Furthermore, SDS-PAGE analysis of the above fractions in the presence of a reducing agent showed that in both fractions a single band was detected at a molecular weight of approximately 20 kDa. The result of the amino acid analyzes of the fractions showed that each of the amino acid compositions was almost in accordance with the corresponding theoretical value assessed from sequence information. In addition, the results of N-terminal amino acid analysis showed that they were the expected sequences. The protein amount of each fraction determined from the results of the amino acid analyzes was 2.56 mg (Fr. G-a) or 1.16 mg (Fr. G-d). After full dialysis of each fraction against IMDM medium, M-07e analysis was performed. As a result, fraction TPO had relative activity of about 3,960,000 (Fr. G-a) or about 7,760,000 (Fr. G-d).

EXAMPLE 62

Construction of recombinant vector pDEF202-hTPO163 for the expression of a partial length human TPO (amino acids 1 to 163) below referred to as "hTP0163" within the CHO cells

The vector pDEF202 constructed in Example 31 was treated with restriction enzymes EcoRI and SpeI followed by isolation of a large vector fragment using agarose gel electrophoresis. The fragment was ligated by T4 DNA ligase (Takara-Shuzo) with an hTPO 163 cDNA that had been generated by symbol processing of a plasmid pEF18S-hTP0163 containing hTP0163 cDNA coding for amino acids-21 to 163 (SEQ ID NO: 13) with EcoRI and SpeI residue symbol to obtain an expression vector pDEF202-hTP0163. This plasmid contained an origin of replication symbol of SV40, a human elongation factor 1 promoter, an SV40 early polyadenylation site, a murine DHFR minigene, an origin of replication of pUC18 and a lactamase gene (Amp<1>), in which the hTP0163 cDNA is linked to the symbol site downstream of the human elongation factor symbol l-a promoter.

EXAMPLE 63

Expression of hTPO 163 in CHO cells

A CHO cell line (dhfr strain, Urlaub and Chasin, Proe. NatLAcad. Sei. USA, 77, p 4216, 1980) was grown in minimal essential medium (α-MEM(-) with thymidine and hypoxanthine) containing 10% fetal bovine serum in a 6 cm dish (Falcon), and then transformed with a pDEF202-hTPO163 plasmid using the tramfectum method (Seikagaku Kogyo K.K.). 10 ng of the plasmid pDEF202-hTPO163 formed in Example 62 was mixed with 240 µl of 0.3 M NaCl and then with esymbol mixture of 20 µl of transfectum and 220 µl of H 2 O. This sample solution was added dropwise to the dish followed by cultivation at 61 in a CO2 incubator. The medium was removed from the dish which was then washed twice with α-MEM(-) one symbol by addition of 10% DMSO containing α-MEM(-) before incubation for 2 minutes at room temperature. Next, 10% dialyzed fetal bovine serum-containing non-selection medium (-MEM(-) with hypoxanthine and thymidine) was added to the dish followed by two days of culture and then selection in 10% dialyzed fetal bovine serum-containing selection medium (a-MEM( -) without hypoxanthine and thymidine). Selection was performed by trypsinizing the cells, dividing these into five 10 cm dishes or twenty 24 well dishes per the above-mentioned 6 cm dish, and continuing cultivation while the medium was changed with fresh selection medium at intervals of two days. The supernatants from the dishes or wells were analyzed for TPO activity using Ba/F3 analysis, whereby the TPO activity was observed. The cells in which TPO activity was observed in the culture supernatants were then transferred to fresh dishes or wells after dividing to a cell concentration of 1:15 with selection medium containing 25 nM methotrexate, and then cultured again for growth and cloning of cells resistant to methotrexate.

Alternatively, transformation of the CHO cells can be performed by co-transfection of pEF18S-hTP0163 and pMG1 into CHO cells.

The CHO strain (CHO-DUKXB11) transfected with plasmid pDEF202-hTPO163 has been deposited by the applicants on January 31, 1995 under accession number FERM BP-4989 to

National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, Ministry of International Trade and Industry, Japan.

EXAMPLE 64

Cultivation of CHO cells symbol large scale

Large-scale cultivation of hTP0163-producing CHO cell line (CHO109 cells, obtained by the above selection in 10% dialyzed bovine fetal serum-containing selection medium [α-MEM" without hypoxanthine and thymidine) which had been obtained by transfection of hTP0163 expression plasmid pDEF202- hTPO163 into the CHO cells of Example 63 was carried out as follows. The CHO 109 cells were cultured in DMEM/F-12 medium (GJBCO) with 10% FCS. After the cells were harvested (or trypsinized) using a trypsin solution, ten million cells inoculated into Falcon roller bottles (Falcon 3000) containing 200 ml of the same medium and then cultured at a roller speed of 1 rpm at 37 °C for 3 days. The culture medium was removed under suction and the surface of the cell culture was rinsed with 100 ml of PBS. 200 ml of DMEM/F-12 medium (GD3CO) without 10% FCS was added to the cell culture followed by 7 days of cultivation at 37 °C at 1 rpm. The harvested culture supernatant was used as a starting material for subsequent e purification step. Similar procedure was performed in 300 roller bottles to provide 601 serum-free culture supernatant.

EXAMPLE 65

Purification of hTPO 163 from the hTP0163-producing CHO cell line

(1) 601 serum-free culture supernatant obtained in Example 64 was filtered through a 0.22 nm filtration filter to collect a filtrate which was then concentrated using an ultrafiltration unit (Filtran, molecular weight 10,000 cut-off) to obtain a concentrated fraction (600 ml, 11.2 mg protein/ml total protein 6430 mg). By Western analyzes of this fraction using anti-HT1 peptide antibody, the presence of hTP0163 protein expressed in a range of apparent molecular weight from 20,000 to 26,000 was shown.

The resulting concentrated culture supernatant (537 ml) was run on a Sephadex G-25 Fine column (Pharmacia-Biotech, Catalog No. 17-0032-02; diameter 10 cm and layer height 30 cm) pre-equilibrated with 10 mM sodium phosphate buffer (pH 6 ,8) to obtain a protein fraction Fl (938 ml, protein concentration 4.9 mg/ml, total protein 4594 mg) as a solution in 10 mM sodium phosphate buffer (pH 6.8).

This protein fraction (929 ml) from the Sephadex G-25 Fine column was applied at a flow rate of 15 ml/min to an SP Sepharose Fast Flow column (Farmacia-Biotech, Catalog No. 17-0729-01; diameter 5 cm and bed height 12 cm) pre-equilibrated with 10 mM sodium phosphate buffer (pH 6.8) followed by elution with 10 mM sodium phosphate buffer (pH 6.8) and then the same buffer containing 10% ethanol. The eluates were combined as fraction F1 (1608 ml, protein concentration 2.13 mg/ml, total protein 3426 mg). Then, the second elution was performed with 10 mM sodium phosphate buffer (pH 6.8) containing 750 mM NaCl and 25% ethanol to collect a major hTP0163 eluate of fraction F2 (651 mL, protein concentration 1.67 mg/mL, total protein 1087 mg) .

To the TPO activity-containing fraction F2 (200 ml) from the SP Sepharose Fast Flow column was added ethanol and purified water to form 300 ml of 45% (final concentration) ethanol solution. Insoluble materials that appeared after the addition of ethanol were removed by centrifugation. The supernatant was injected into the SOURCE 15RPC column

(Pharmacia-Biotech, Catalog No. 17-0727-02; diameter 2 cm and layer height 20 cm)

pre-equilibrated with 50% solvent A (10 mM sodium acetate buffer, pH 6.7) plus 50% solvent B (10 mM sodium acetate buffer, pH 6.7, containing 90% ethanol) at a flow rate of 2 ml/min. Then the column was washed with 55% solvent A plus 45% solvent B until the unabsorbed compounds were almost completely eluted, after which elution at a flow rate of 1.5 ml/min was performed according to the following elution profile: 50% B for 5 minutes; linear gradient from 50% B to 100% B over 140 min and then 100% B for 35 min. The fractions were collected at intervals of 5 minutes (equivalent to 7.5 ml volume). All the fractions were subjected to SDS-PAGE and then to Western analyzes to investigate an elution range of hTPO 163. As a result, highly purified hTPO 163 with an apparent molecular weight from about 20,000 to 26,000 was found to be eluted in a range from 66% to 87% ethanol. Among these hTP0163 proteins, hTP0163 with a higher molecular weight was eluted earlier from the reverse phase column and this suggests that the glycosylated hTPO 163 molecule has an increased hydrophilicity.

TII 88.8 ml hTP0163 elution fraction (ie hTP0163 fraction (90ml) eluted in a range from 68% to 86.5% ethanol) from SOURCE 15RPC column was added CHAPS and the mixture was concentrated and washed to obtain 2.5 ml of a concentrate containing approximately 5% ethanol and 4 mM CHAPS. This concentrate was then applied to a Superdex 75 pg column (Pharmacia-Biotech, Catalog No. 17-1070-01; diameter 2.6 cm and layer height 60 cm) equilibrated with 10 mM sodium phosphate buffer (pH 6.8) containing 10% ethanol at a flow rate of 1.5 ml/min. From the time point 60 min after application, elution fractions were collected in an amount of 6 ml each (ie every 4 min). As a result, hTP0163 was eluted in fractions from tube No. 16 at least tube No. 31 as determined by SDS-PAGE. This elution position corresponds to a molecular weight range from about 44,000 to about 6,000 as determined by gel filtration using a standard molecular weight marker (mixture of Bio-Rad, Gel Filtration Standard, Catalog No. 151-1901; and Calbiochem, insulin, Catalog No. 407696). In addition, these hTPO 163 molecules appeared to be eluted in a series of decreasing degrees of glycosylation. All hTPO 163 species in the range from tube numbers 16 to 31 (elution volume: 108 to 276 ml) were collected as fraction FA. In addition to fraction FA, the fractions with tube numbers 16 to 18 (elution volume: 180 to 198 ml), 19 to 24 (elution volume: 198 to 234 ml) and 25 to 31 (elution volume: 234 to 276 ml) were separately collected and designated as fractions FH, FM and FL. Fraction FA was further formed by combining part of the fractions FH, FM and FL. The above is shown in fig 19.

(2) Next, the hTP0163 elution fractions FH, FM, FL and FA from the Superdex 75 pg column indicated in (1) will be illustrated in further detail. The N-terminal amino acid sequences of the hTPO 163 species obtained above were determined according to the same method as in Example 1, but most of the N-terminal Ser residues could not be identified. This suggests that an O-linked sugar compound is added to N-terminal Ser. The sequence after the N-terminal Ser was confirmed to be an amino acid sequence extracted from the gene sequence of hTPO. The concentrations of the hTP0163 proteins in FH, FM, FL and FA were 10.2 ng/ml, 6.2 ng/ml, 0.84 ng/ml and 3.2 ng/ml as determined by amino acid analysis (AccQ. Tag method, Wasters), provided that the concentrations had peptide moieties that did not contain a sugar chain. These fractions (100 ng each) were subjected to SDS-PAGE under non-reducing conditions using a multigel 15/25 (Dai-ichi Kagaku

Yakuhin, 15 to 25% precast polyacrylamide gel) or under reducing conditions using DDT, followed by silver staining (Daiichi Pure Chemicals). As a result, each of these fractions was found to contain highly pure hTPO 163. Apparent molecular weights of hTPO 163 in FH, FM, FL and FA, as calculated using DPCin molecular weight marker (Daiichi Pure Chemicals) as a standard under reducing conditions, were 24,000 to 21,500, 23,000 to 21,000, 23,000 to 20,500 and 23,500 to 20,500 (see fig 20). In addition, apparent molecular weights of hTPO were FH, FM, FL and FA, calculated using a biotin-labeled SDS-PAGE standard (Bio-Rad, Biotynylated SDS-PAGE Standards, Broad Range: Catalog No. 161-0319) under reducing conditions on Western analysis, 26,000 to 22,000, 25,000 to 22,000, 26,000 to 21,000 and 26,000 to 21,000. The heterogeneity among the molecular weights of FH, FM, FLog FA may be caused by the heterogeneity of O-linking sugar chains. Each fraction of FH, FM, FL and FA was digested with neuraminidase (Neuraminidase, Nacalai tesque, Catalog No.242-29SP) reduced with DTT and then analyzed on SDS-PAGE. As a result, the molecular weights were around 19,000 in all the fractions and this suggests that the heterogeneity of the molecular weight of hTPO is mainly caused by the heterogeneity of the amount of sialic acid in the sugar chains linked with the hTPO 163 protein and that the hTP0163 obtained was expressed as a glycoprotein in the CHO cells. (3) The fractions FH, FM, FL and FA obtained in (2) were analyzed for in vitro activity according to the M-07e analysis system. As a result, the relative specific activity was 511,000,000, 775,000,000, 1,150,000,000 and 715,000,000/mg hTP0163 protein (weight of peptide portion not containing sugar weight).

EXAMPLE 66

Construction of E.coli vector for expression of human TPO (amino acids 1 to 332) where Lys is added in position-1 and Met in position-2 (referred to below as "hMKT(1-332)") and expression of hMKT (l-332)

To express the full-length amino acid sequence of human TPO in E.coli, inverted codons from amino acid 164 to amino acid 332 were replaced with E.coli preference codons as described below.

Synthetic oligonucleotides: 21 and 22; 23 and 24; 27 and 28; 29 and 30; 21 and 32; 33 and 34; 35 and 36; 37 and 38; or 39 and 40 were phosphorylated in a solution of 0.1 mM ATP, 10 mM Tris-acetate, 100 mM Mg-acetate, 50 mM K-acetate using T4 kinase (Pharmacia) in the same tube. Then 1/10 volume of a solution of 100 mm Tris/HCl (pH 7.5), 100 mm MgCl 2 , 500 mm NaCl was added to the reaction mixture, boiled for 3 minutes in a water bath and cooled to form a double-stranded DNA. Four sets of double-stranded DNA, ie oligonucleotides 21 and 22/23 and 24 (combination-A); oligonucleotides 25 and 26/27 and 28 (combination-B); oligonucleotides 31 and 32/33 and 34 (combination-C) and oligonucleotides 35 and 36/37 and 38/39 and 40 (combination-D) were separately ligated with a DNA ligation assay kit (Takara-Shuzo), and then combination- A ligated together with combination-B and likewise combination-C with combination-D to form ligation-1 and ligation-2 respectively. When using ligations -1 and -2 as templates, PCR was carried out using nucleotides 41 and -42 for ligation-1, as well as oligonucleotides-43 and -44 for ligation-2, as primers. The PCR products of ligations-1 and -2 were digested with SacI and EcoRV and with EcoRV and Hindin, followed by electrophoresis on 2% agarose gel and then purification using a Prep-A-Gene DNA purification assay kit to isolate approximately 240 bp and 250 bp fragments. Two fragments obtained in this way were subcloned into pBluescript IIKS+ (Strategene) previously digested with SacI and Hindin (E.coli DH5 was used as a host). Of the resulting clones, the clone having the base sequence shown in Table 6 was selected for sequencing and designated pBL(SH)(174-332) (see SEQ ID NO: 154).

Subsequently, pBL(XH)h6T(1-163) was generated in Example 42 as a template subjected to PCR in the presence of the following synthetic oligonucleotides 45 and 46 as primers, and PCR products were digested with BamHI and SacI followed by electrophoresis on a 6% polyacrylamide gel for to isolate a fragment of approximately 160 bp from the gel.

Furthermore, pBL(SH)(174-332) was digested with SacI and Hindin and the digest was then purified using the Prep-A-Gene DNA Purification assay kit to provide a fragment of approximately 480 bp. The two fragments generated above were subcloned into pBluescript IIKS+ (Stratagene) previously digested with BamHI and Hindin (E.coli DH5 was used as a host).

Of the resulting clones, the clone having the base sequence shown in Table 7 was selected by sequencing and designated pBL(BH)(123-332) (seeSEQ ID NO: 157). pBL(XH)h6T(l-163) prepared in Example 42 was digested with BamHI and HindDI, and pBL(BH)(l23-332) prepared above was likewise digested to produce a fragment of about 640 bp after the two fragments were ligated together to obtain a clone pBL(XH)h6T(1-334). PCFM536/hMKT(1-163) prepared in Example 52 was digested with XbaI and SfiI to provide an approximately 270 bp fragment then ligated with pBL(XH)hMKT(1-334) digested with the same restriction enzymes to produce a clone pBL (XH)hMKT(1-334). After pBL(XH)hMKT(1-334) was cleaved with XbaI and Hindin, an approximately 1040 bp fragment encoding mutation-type human TPO amino acids 1-334 was isolated and purified, and then cloned into pCFM536 (EP-A-136490) pre-cleaved with XbaI and Hindin (E.coli JM109 pre-transformed with pMW1 (ATCC No. 39933) used as a host). The resulting clone was designated pCFM536/- hMKT(l-332) and the E.coli strain containing this expression vector was used as a transformant for expression of mutant type protein hMKT(l-332). Expression plasmid pCFM536/hMKT(1-332) contains a DNA sequence shown in SEQ ID NO: 14.

The above transformant was grown in 60 ml LB medium containing 50 ng/ml ampicillin and 12.5 ng/ml tetracycline overnight at 30 °C, and the culture (25 ml) was then added to 1000 ml LB medium containing 50 ng/ml ampicillin followed by shaking of the culture at 36 °C until the ODgQO reached 1 0 nl 1 -2- After approximately 330 ml of LB medium at 65 °C was added to the culture so that the final temperature of the culture was 42 °C, the shaking culture was continued for a further 3 hours at 42 °C to induce expression of variant human TPO, hMKT(1-332) protein, also referred to as [Met"<2>, Lys-^TPOØ-332).

The obtained culture was directly subjected to SDS-PAGE analysis. In the SDS-PAGE analysis, Multigel 15/25 (Dai-ichi Chemical Co) was used. After electrophoresis and Coomassie Blue staining, protein specific for induction of expression was detected at a molecular weight of approximately 35 KD on the gel with respect to the transformant so that the expression of hMKT(1-332) protein was induced. In addition, after SDS-PAGE and electroblotting (using nitrocellulose membrane), the above expressed protein was reacted with anti-HT1 peptide antibody formed in Example 45. After staining, a band was detected at a molecular weight of about 35 kD which consequently indicates that hMKT(l- 332) was expressed.

EXAMPLE 67

Preparation of human TPO substitution derivatives, expression in COS7 cells and identification of activity

According to the following procedures, if several derivatives of the amino acids for human TPO were partially substituted with other amino acids, the TPO activity has been studied.

Vector pSMT201 for expression in animal cells, for use in this example, was constructed as follows.

First, expression plasmid pCM-mEPORn containing a murine erythropoietin receptor cDNA (obtained from Dr D'Andrea in the Dana Farbor Cancer Institute; Cell: 57,277-285 (1989)) was treated with the restriction enzymes KpnI and EcoRI and then electrophoresed on agarose gel to isolate a pXM vector fragment in which the murine erythropoietin receptor cDNA had been removed. Then, two oligonucleotides for introducing different cloning restriction sites into the above expression vector were synthesized using ay DNA synthesizer (ABI). The oligonucleotides synthesized had the following base sequence:

The two oligonucleotides were mixed and cooled to form a double nucleotide which was then ligated with the pXM vector obtained above in the presence of T4 DNA ligase (Takara-Shuzo) to give an expression vector pDMT201. To remove the murine DHFR cDMA contained in pDMT201 from pDMT201, the vectors pDMT201 and pEF18S were separately digested with Noti and HpaI followed by agarose gel electrophoresis to isolate a large fragment from pDMT201 vector DNA and a smaller fragment from pEF18S vector DNA. Then, these fragments were ligated together using T4 DNA ligase (Takara-Shuzo) to produce an expression vector pSMT201. pSMT201 as shown in Fig. 21 has an origin of replication of SV40, an enhancer sequence, adenovirus major late promoter sequence, an adenovirus tripatite leader sequence, a splice signal sequence, an SV 40 early polyadenylation site sequence, an adenovirus VA RNA gene sequence, an origin of replication of pUC18 and a β-lactamase gene ( Amp<1>), along with the following restriction sites for ligation of a desired gene: BgHI, PstI, KpnI, XhoI, EcoRI, SmaL SpeI and Noti sites. The pHTP illustrated in Example 30 containing full-length human TPO cDNA was digested with EcoRI and SpeI to provide a full-length human TPO cDNA fragment which was then subcloned into a vector pSMT201 previously cleaved to produce a plasmid pSMT201-hTPO.

Using the thus obtained plasmid pSMT201-hTPO as a template, plasmid pGL-TPO/pBlue which was used as a template for the preparation of the derivatives of interest was first constructed as follows.

In pGL-TPO/pBlue, addition of the rabbit β-globin leader sequence to the S^ non-translational site of human TPO was attempted according to Annweiler's method (Annweiler et al, Nucleic Acids Research, 19: 3750, 1991). For this purpose, the following two DNA strands were chemically synthesized:

ligated with a fragment obtained by digesting the vector Bluescript II SK+ (Toyo-boseki) with EcoRI and Smal to produce a vector pGL/pBlue in which the leader sequence was inserted.

PCR was performed using the following primer sequences:

(ie, 102-122 sequence of SEQ ID NO. 7 to which the Smal sequence had been ligated); and

(ie, antisense to the 1140-1160 sequence of SEQ ID NO:7, to which the Spel and HindTf1 sequences have been ligated).

PCR was performed using 1 ng of the plasmid pSMT201-hTPO DNA as a template together with 10 n m synthetic primers. Using the TaKaRa PCR Amplification Kit (Takara-Shuzo) and Programmable Thermal Controller (MJ Research), PCR was performed in a volume of 100 ni under the following conditions: 3 cycles of 94 °C 1 min, 55 °C 2 min and 72 °C C 2 min/cycle; then 20 cycles of 94 °C 45 sec, 55 °C 1 min and 72 °C 1 min/cycle. The PCR product was extracted with chloroform and then methanol-precipitated twice followed by suspension in 100 nI TE buffer.

The PCR product was then subjected to restriction cleavage with Smal and Hindin, extracted with phenol/chloroform and precipitated with ethanol. The resulting precipitate was dissolved in 10 µl TE buffer and then subcloned into vector pGEL/pBlue which had been previously digested with the same restriction enzymes (Competent.High E.coli JM109 (Toyo-boseki) was used as a host). Of the transformant cells obtained, 20 clones were selected from which plasmid DNA was generated essentially as described in Sambook et al, Molecular Cloning, Cold Spring Harbor Laboratory Press (1989). Purified plasmid DNA was identified by sequencing using the Taq Dye Deoxy™ Terminaler Cycle Sequencing Kit (Applied Biosystems, Inc) and 373A DNA sequencer (Applied Biosystems Inc). As a result, it was confirmed that the plasmid encoding pGL-TPO/pBlue had the expected TPO cDNA sequence without any substitution over the entire length. After pGL-TPO/pBlue was cleaved with Bgffl and Spel, the cleavage was extracted with phenol/chloroform and then ethanol-precipitated. The obtained precipitate was dissolved in 10 ni TE buffer and was subcloned into vector pSMT201 which had been digested with the same enzymes (CompetentHigh E.coli JM109 (Toyo-boseki) was used as a host). According to the method described in Molecular Cloning (Sambrook et al, supra), plasmid DNA was generated from transformant cells. The resulting expression vector pSMT/pGL-TPO had a structure where the p-globin leader sequence and human TPO cDNA were linked at the BglH/Spel restriction site downstream of the "sprice" signal sequence of the pSMT201 vector as shown in Fig. 21. This pSMT/pGL-TPO vector was used for transfection into COS cells.

For the formation of human TPO substitution derivatives, PCR was used where the method of Ito was used (Ito et al, Gene, 102:67-70, 1991). Two derivatives of human TPO in which Arg-25 and His-33 of human TPO are replaced by Asn and Thr were generated for illustration.

These derivatives may be referred to as [Asn<25>]TPO and [Thr<33>]TPO. In the PCR, the following primers were used:

T7: 5-TAATACGACTCACTATAGGGCG-31 (SEQ ID NO: 164)

(corresponds to the T7 promoter region of Bluescript IISK+);

ABglH: 5'-AATTCCAAGATCACACACTTGC-3' (SEQ ID NO: 165)

(for substitution of the BgHI recognition sequence);

end: 5'-TCAAGCTTACTAGTCCCTTCCTGAGACAGATTCTG-3' (SEQ ID NO: 166)

(antisense to 1140-1160 of SEQ ID NO: 7 to which SpeI and HindUJ sequences have been ligated);

N3: 5-TGGGCACTGGCTCAGGTTGCTGTGAAGGACATGGG-3' (SEQ ID NO:

167)

(Arg 25 -> Asn to SEQ ID NO: 7); and

09: 5-TGTAGGCAAAGGGGTAACCTCTGGGCA-31 (SEQ ID NO: 168)

(His-33 -> Thr of SEQ ID NO: 7).

First step PCR was performed using 1 ng of plasmid pVGL-TPO/pBlue DNA as a template together with 10 ng of each of the synthetic primers. Combinations of the primers were: [1] ABgUI and the "end" primers; [2] T7 and N3; and [3] T7 and 09. Using the TaKaRa PCR Amplification Kit (Takara-Shuzo) and Programmable Thermal Controller (MJ Research), PCR was performed in a volume of 100 ni under the following conditions: 3 cycles of 94 °C 1 min, 55 °C 2 min and 72 °C 2 min/cycle; then 17 cycles of 94 °C 1 min, 55 °C 2 min and 72 °C 2 min/cycle. Each of the PCR products was extracted with chloroform and precipitated twice with ethanol, and the precipitate was dissolved in 100 ni TE buffer. The resulting PCR products (1 nine each) were then subjected to another PCR. The combinations of the templates were: PCR products of [l]/[2] and PCR products of [l]/[3]. With regard to the primers, the combination of T7 and end was used in two cases. After incubation at 94 °C for 10 min and addition of TaKaRa Taq, PCR was performed under the following conditions: 3 cycles at 94 °C 1 min, 55 °C 2 min and 72 °C 2 min/cycle; then 9 cycles of 94 °C 45 sec, 45 °C 1 min and 72 °C 1 min/cycle. To each of the resulting PCR products was added 1 µl proteinase K (5 mg/ml) 2 µl 0.5 M EDTA and 2 µl 20% SDS, incubated at 37 °C for 30 min to inactivate Taq, extracted with phenol /chloroform and ethanol-precipitated. Then the precipitate was dissolved in sterile water and digested with BgJJI and Spel, followed by phenol/chloroform extraction and ethanol precipitation. The precipitate obtained in this way was dissolved in 10 ni TE buffer and then subcloned into the expression vector pSMT201 which had previously been digested with the same restriction enzymes and treated with calf intestinal alkaline phosphatase (Boehringer-Mannheim) (Competent High E.coli: JM109 was used as a host). Of the transformant cells obtained, in each case two clones were selected from which plasmid DNA was formed essentially as described in Molecular Cloning (Sambook et al, supra). Purified plasmid DNA was sequenced using the 373A DNA sequencer and the Taq Dye Deoxy™ Terminaler Cycle Sequencing Kit (both Applied Biosystems Inc.).

Plasmid formed by PCR using the primers [1] and [3] contained a cDNA coding for a TPO substitution derivative (09/TPO). The amino acid substitution of His33(CAC) with Thr(ACC) and extension of the original C-terminus (Gly 332) by addition of the amino acid sequence TSIGYPYDWDYAGVHHHHHH (SEQ ID NO: 169) was confirmed. This derivative was referred to as [Thr<33>, Thr333, Ser334, Ile335, Gly336, Tyr33? Pro338, Tyr33?, Asp340, Val341, Pro342, Asp343, Tyr344, Ala345, Gly346, Val347, His348, His34?, His35<0>, His35<1>, His3<52>, His<353>]TPO.

Plasmid formed according to PCR using the primers [1] and [2], on the other hand, contained a cDNA coding for a TPO substitution derivative (N3/TPO). The amino acid substitutions of Arg25(AGA) with Asn(AAC) and Glu231(GAA) with Lys(AAA) and extension of the original C-terminus (Gly332) by adding the amino acid sequence TSIGYPYDWDYAGVHHHHHH were confirmed. This derivative was referred to as [Asn25, Lys231, Thr333, Ser334, Ile335, Gly336, Tyr337, Pro338, Tyr33? Asp340, Val34<*>, Pro342, Asp343, Tyr34<4>, Ala345, Gly346, Val34<7>, His348, His34? His350, His351, His35<2>, His<353>]TPO.

Transfection into the COS7 cells of each clone obtained was performed according to the DEAE-dextran method including chloroquine treatment as described in Example 35. 40 ng of each of the plasmid DNAs was used in the transfection and after 5 days the culture supernatant was isolated and then concentrated to a volume of 1/20 using Centricon-30 (Amicon) to assess their TPO activity according to the M-07e assay. As a result, the culture supernatants of the COS7 cells into which plasmids containing cDNA encoding the human TPO substitution derivatives N3/TPO and 09/TPO were transfected, the TPO activity was detected in a dose-dependent manner (see Fig. 22).

EXAMPLE 68

Preparation of insertion or deletion derivatives of human TPO, expression thereof in COS7 cells and identification of TPO activity

This example shows whether insertions or deletion derivatives of hTPO 163 have a TPO activity.

The derivatives formed were a His33 deletion derivative ([AHis<33>]TPO(1-163), dH33); a Glyl 16 deletion derivative ([AGly<116>]TPO(1-163), dGl 16); an Arg1 17 deletion derivative ([AArg<117>]TPO(1-163), dR116), a Thr insertion derivative ([His<33>, Thr33', Pro34]TPO-(1-163), T33'), an Ala insert derivative ([His<33>, Ala<33>', Pro<34>]TPO(l-163), A33') and a Gly insert derivative ([His<33>, Gly33', Pro34, Ser <38>]TPO(1-163), G33) in which Thr, Ala and Gly have been independently inserted between His33 and Pro34; and an Asn insert derivative ([Gly116, Asn116', Arg117]TPO(l-163), NI 16'), an Ala insert derivative ([Gly116, Ala116', Arg117]TPO(l-163), Al 16') and a Gly insertion derivative ([Gly116, Gly116', Arg<117>]TPO(l-163), Gl 16') in which Asn, Ala, and Gly have been independently inserted between Glyl 16 and Argl 17, and all sequences are based on SEQ ID NO: 13.

Among the derivatives, T331 and NI 16' are thought to introduce mucin-type and Asn-bond type sugar chains.

The above derivatives were formed using PCR according to the method described by Ito et al (Gene, 102: 67-70, 1991). The primer sequences used in PCR are as follows: dH33: 5'-TGTAGGCAAAGGAACCTCTGGGCA-3' (SEQ ID NO: 172) (for preparation of the His33 deletion derivative based on the sequence shown in SEQ ID NO: 7); T33': 5'-TGTAGGCAAAGGAGTGTGAACCTCTGG-3' (SEQ ID NO: 173) (for the preparation of the Thr insert derivative where Thr is inserted between His33 and Pro34 in the sequence shown in SEQ ID NO: 7);

A33': 5-TGTAGGCAAAGGAGCGTGAACCTCTGG-31 (SEQ ID NO: 174) (for the preparation of the Ala insert derivative where Ala is inserted between His33 and Pro34 in the sequence shown in SEQ ID NO: 7);

G33': 5,-TGTAGGCAAAGGTCCGTGAACCTCTGG-3' (SEQ ID NO: 175) (for the preparation of the Gly insert derivative where Gly is inserted between His33 and Pro34 in the sequence shown in SEQ ID NO: 7);

dGl 16: 5'-AGCTGTGGTCCTCTGTGGAGGAAG-3' (SEQ ID NO: 176) (for preparation of the Glyl 16 deletion derivative based on the sequence shown in SEQ ID NO: 7); dRl 17: 5-GTGAGCTGTGGTGCCCTGTGGAGG-3* (SEQ ID NO: 117) (for the preparation of the Arg1 17 deletion derivative based on the sequence shown in SEQ ID NO: 7); NI 16': S^AGCTGTGGTCCTGTTGCCCTGTGGAGG-S' (SEQ ID NO: 178) (for the preparation of the Asn insert derivative where Asn is inserted between Glyl 16 and Argl 17 in the sequence shown in SEQ ID NO: 7);

Al 16': 5-AGCTGTGGTCCTAGCGCCCTGTGGAGG-3' (SEQ ID NO: 179) (for the preparation of the Ala insert derivative where Ala is inserted between Glyl 16 and Argl 17 in the sequence shown in SEQ ID NO: 7);

Gl 16': 5'-AGCTGTGGTCCTTCGCCCTGTGGAGG-3* (SEQ ID NO: 180) (for the preparation of the Gly insert derivative where Gly is inserted between Glyl 16 and Argl 17 1 sequence shown in SEQ ID NO: 7);

dRI: 5-CGGGCTGCAGGATATCCAAGATCTCA-31 (SEQ ID NO: 181) (for substitution of a restriction enzyme EcoRI recognition sequence);

T7: 5'-TAATACGACTCACTATAGGGCG-3' (SEQ ID NO: 182) (corresponds to the T7 promoter region of BluescriptnSK<4>"); and endNotl: 5'-GGGCGGCCGCTCAGCTGGGGACAGCTGTGGTGGG-3' (SEQ ID NO: 183)

(antisense to the 633-653 nucleotide sequence of SEQ ID NO: 7, where a termination codon and a Noti recognition sequence have been added).

A first step in PCR was performed using, as a template, 1 ng of plasmid pGL-TPO/pBlue DNA generated in Example 67 and using 10 µg of synthetic primers. The combination of primers was: [1] dRI and endNot1 or [2] T7 and mutated primers (dH33, A33', G33', T33\ dGl 16, dRI 17, Al 16', Gl 16' and NI 16'). In PCR, TaKaRa PCR Amplification Kit (Takara-Shuzo) and Programmable Thermal Controller (MJ Research) were used and PCR was performed in a final volume of 100 µl. In the first PCR reaction according to [1], one cycle was 94 °C 1 min, 45 °C 2 min and 72 °C 2 min and 3 cycles were performed, followed by 17 cycles where one cycle was 94 °C 45 sec, 45 ° C 1 min and 72 °C 1 min. In contrast, the first PCR reaction according to [2] was performed under conditions of 94 °C 1 min, 55 °C 2 min and 72 °C 2 min for 3 cycles followed by 94 °C 45 sec, 55 °C 1 min and 72 °C C 1 min for 17 cycles. Each of the PCR products was extracted with chloroform before two ethanol precipitations, and then the resulting precipitate was dissolved in 100 µl TE buffer.

Then another PCR step was performed using 1 ni of each of the PCR products according to [1] and [2]. The primers used were a combination of T7 and EndNotI. After incubation at 94 °C for 10 min, TaKaRa Taq was added to each PCR reaction mixture. The second PCR was performed under conditions of 94 °C 1 min, 55 °C 2 min and 72 °C 2 min for 3 cycles followed by 94 °C 45 sec, 55 °C 1 min and 72 °C 1 min for 9 cycles . To each PCR product obtained in this way, 1 µl of 5 mg/ml proteinase K, 2 µl of 0.5 M EDTA and 2 µl of 20% SDS were added, and the mixture was kept at a temperature of 37 °C for 30 min to inactivate Taq. The PCR product was extracted with phenol/chloroform followed by ethanol precipitation, and the resulting precipitate was then dissolved in 20 µL of sterile water. After digestion with EcoRI and Noti, the solution was extracted with phenol/chloroform and then precipitated with ethanol. The obtained precipitate was dissolved in 10 ni TE buffer, and subcloned into an expression vector pEF18S which had previously been cleaved with the same restriction enzymes and treated with alkaline phosphatase (from calf intestine, Boehringer-Mannheim). A host cell used was Competent-High E.coli JM109 (Toyo-Boseki). From the transformant thus obtained, a plasmid DNA was generated substantially according to the method described in Molecular Cloning (Sambook et al, Cold Spring Harbor Laboratory Press (1989)). Then became

the nucleotide sequence of purified plasmid DNA confirmed using Taq Dye

in the Deoxy™ Terminaler Cycle Sequencing Kit (Applied Biosystems) and using the Applied Biosystems 373A DNA sequencer. As a result, it was confirmed that all the plasmids encoding dH33, A33', T33', dGl 16, dRI 17, Al 16', Gl 16' and NI 16' had expected TPO cDNA sequence without any substitution of a nucleotide sequence by

seats different from the assumed seats. IG33', a base substitution [Pro38 (CCT) Ser (TCT)] different from that at the putative site was observed.

Transfection of each clone obtained into COS7 cells was carried out according to the method of Example 11. The transfection was carried out using 10 ng of each plasmid DNA

according to the DEAE-dextran method including chloroquine treatment, and after 4 to 5 days in the culture supernatant was isolated and then assessed according to the M-07e analysis system. As a result, TPO activity was detected in each supernatant of the COS7 cell cultures transfected with plasmids encoding an arninoic acid insertion or a deletion derivative in a dose-dependent manner (see Fig. 23). In contrast, no TPO activity was found in

the supernatant of the COS7 cell culture showing that expression plasmid pEF18S does not

i contained a cDNA coding for a TPO derivative that had been transfected in and expressed.

EXAMPLE 69

Preparation and purification of anti-human TPO peptide antibodies

(1)6 regions (shown in Table 8) which are considered relatively suitable as antigens were

selected from the amino acid sequence of human TPO according to SEQ ID NO: 191, from which a 4-stranded peptide of a multiple antigen peptide (MAP) type was synthesized according to Tams

method (Proe. Nati. Acad. Sei., USA, 85,5409-5413,1988). Two rabbits were immunized eight times each with 100 µg of this peptide.

Single-stranded peptides were produced by ligation of cysteine residues to the C-terminus of each of the peptide regions shown according to SEQ ID NOS: 119-124 (these are referred to as

peptide region HT 1 to 6 regions, which are partial peptides corresponding to 8 to 28, 47 to 62, 108 to 126, 172 to 190, 262 to 284 and 306 to 332, of the sequence according to SEQ ID NO: 191). When using these as test antigens, the antibody values in the sera obtained from the rabbits immunized in this way were measured according to enzyme immunoassay from which the increase in the antibody value was confirmed in all sera that were tested. These sera were consequently referred to as anti-sera.

In addition, the above-mentioned single-stranded synthetic peptide with the cysteine residue attached to its C-terminus was also used as an antigen for affinity purification of anti-sera in the following (2). (2) Two rabbits were each immunized with one of the above-mentioned antigenic peptides and the antibody derived from each of all rabbits thus immunized was raised separately. For anti-HT1 peptide antibodies, anti-HT1-1 peptide antibody and anti-HT1-2 peptide antibody derived from two immunized rabbits were obtained.

As an example, purification of anti-HT1-1 peptide antibody is illustrated below.

First, 30 mg of single-stranded peptide of HT1 with a cysteine residue attached thereto was coupled to 12 ml of Sulfo Link Coupling Gel (manufactured by Pierce Co, according to catalog No. 44895). A peptide solution containing the antigen was coupled to the gel that had been equilibrated with coupling buffer (50 mm Tris, 5 mm EDTA-Na, pH 8.5) at 6 times the volume of the gel, over a period of 15 minutes. After being left static as it was for 30 minutes, the gel was washed with coupling buffer three times the volume of the gel. Then, coupling buffer containing 0.05 M L-cysteine-HCl was added to the gel at a rate of 1 ml/ml gel whereby untreated groups were blocked over a period of 15 minutes. After being left for 30 minutes, the gel was washed with coupling buffer at 8 times the volume of the gel. The above coupling operation was carried out at room temperature. In this way, the antigen region-containing peptide was coupled on the gel by covalent binding, at a coupling efficiency of 28.3% to form an antigen peptide gel with 0.8 mg of the peptide coupled on 1 ml of gel, in the antigen column. After collecting all the blood from each immunized rabbit, 78.4 ml of anti-serum containing anti-HT1-1 peptide antibody was obtained. 76.7 ml of anti-serum (with a protein content of 3620 mg) was added to the antigen column which had previously been equilibrated with 50 mM phosphate buffer (pH 8) containing 150 mM NaCl and 0.05% sodium azide and then the column was washed with the same buffer. Consequently, 105.9 ml of a passing fraction (with a protein content of 3608 mg) was obtained. Then the adsorbed fraction was eluted with 0.1 M citric acid buffer (pH 3.0) which was then immediately neutralized by adding 21.1 ml of 0.1 M carbonate buffer (pH 9.9), concentrated by ultrafiltration (using YM30 membrane produced by Amicon Co.) and purified in 50 mM phosphate buffer (pH 8) containing 150 mM NaCl and 0.05% sodium azide. Accordingly, 11.2 ml of a purified anti-HT1-1 peptide antibody (with a protein content of 77.7 mg) was obtained in the buffer. In the same manner as above, anti-HT1-2 peptide antibody (60.0 mg), anti-HT2-1 peptide antibody (18.8 mg), anti-HT2-2 peptide antibody (8.2 mg), etc. were obtained.

Anti-HT3 to anti-HT6 peptide antibodies are obtained in the same manner as above.

EXAMPLE 70

Detection of TPO Western analyses

(1) To evaluate the anti-sera obtained in Example 69, these were tested in the same manner as mentioned below.

A standard sample of recombinant human TPO which had been partially purified from the supernatant of the culture of CHO cells into which a gene encoding amino acids -21 to 332 of SEQ ID NO:6 had been introduced and expressed was subjected to SDS-polyacrylamide electrophoresis (SDS-PAGE ) according to a common method and then electroblotted onto a PVDF or nitrocellulose membrane. After the blotting, the membrane was washed with 20 mM Tris-HCl and 0.5 M NaCl (pH 7.5) (TBS) for 5 min, then washed twice each with 0.1% Tween 20 containing TBS (TTBS) for 5 min and treated with a blocking agent ( Block Ace; manufactured by Dai-Nippon Pharmacutical Co, as Catalog No. UK-B25) for 60 minutes. Anti-sera each containing one of anti-HT1-1 peptide antibody, anti-HT2-1 peptide antibody, anti-HT3-2 peptide antibody, anti-HT4-1 peptide antibody, anti-HT5-2 peptide antibody and anti-HT6-1 peptide antibody were each diluted with TTBS solution containing 0.05% BSA and 10% Block Ace to 1/1000 dilutions. The diluted antibodies were used for Western analysis as primary antibodies. Recombinant human TPO sample processed as above was treated with TTBS solution containing anti-TPO peptide antibody, 0.05% BSA and 10% Block Ace for 60 minutes and then washed with TTBS for 5 minutes. This was then treated with TTBS solution containing anti-rabbit (ab') goat biotinylated antibody (manufactured by Caltag Co. as catalog No. L43015), as secondary antibody, together with 0.05% BSA and 10% Block Ace, for 60 minutes and then washed twice each with TTBS for 5 min. This was then treated with a 1/5000 dilution of alkaline phosphatase-labeled avidin (manufactured by Leinco Technologies Co. as Catalog No. Al 08) diluted with 10% Block Ace-containing TTBS solution, for 30 minutes, and then washed twice each with TTBS for 5 min and then with TBS for 5 min. Subsequently, this was developed using an alkaline phosphatase substrate (manufactured by Bio-Rad Co., as Catalog No. 170-6432). The above Western analysis was performed at room temperature which verified that human TPO was recognized and detected by the anti-sera. (2) 3 mg of each of anti-HT1-1 peptide antibody, anti-HT1-2 peptide antibody, anti-HT2-1 peptide antibody and anti-HT2-2 peptide antibody which had been purified by affinity chromatography in Example 69 were weighed and biotinylated by coupling of these on activated biotin (NHS-LC-Biotin II, manufactured by Pierce Co, as Catalog No. 21336). The same standard sample of recombinant human TPO as used in the foregoing (1) was subjected to SDS-PAGE, then electroblotted onto a PVDF or nitrocellulose membrane, and subjected to conventional Western analysis using each of these biotinylated antibodies as primary antibodies. Immediately after blotting, the membrane was washed with TBS for 5 min and then washed twice each with TTBS for 5 min, and then this was treated with a blocking agent (Block Ace) for 60 min. This was then treated with TTBS solution containing 1 ng/ml biotinylated anti-TPO peptide antibody, 0.05% BSA and 10% Block Ace for 60 minutes and then washed twice each with TTBS for 5 minutes. This was then treated with a 1/5000 dilution of alkaline phosphatase-labeled avidin (manufactured by Leinco Technologies Co. as Catalog No. A108) diluted with 10% Block Ace-containing TTBS solution, for 30 minutes, and then washed twice each with TTBS for 5 minutes and then TBS for 5 minutes. This was then developed using an alkaline phosphatase substrate

(manufactured by Bio-Rad Co, as Catalog No. 170-6432). The above Western analysis was performed at room temperature which verified that human TPO was recognized and detected by purified antibodies.

EXAMPLE 71

Preparation of anti-TPO peptide antibody columns and detection of human TPO

in

Human TPO peptide antibodies obtained in Example 69 were verified to recognize human TPO. These antibodies were separately coupled onto a chromatography support to form anti-TPO peptide antibody columns. As an example, preparation of anti-HT1-2 peptide antibody columns is illustrated below.

(1) A solution comprising 50 mM Na phosphate and 0.15 M NaCl (pH 8.0) and containing 5 mg/ml antibody was formed. 0.8 ml of this solution was combined and coupled with 1.54 ml of a swollen formyl-activated gel (Formyl-Gellulofine, manufactured by

Chisso Co.) at 4 °C for 2 h. Then, 1.1 ml of a solution containing 10 mg/liter of a reducing agent (trimethylamine borane (TMAB) produced by Seikagaku Kogyo K.K as catalog No. 680246) was added and further coupled for another 4 hours. Then, only the gel portion was isolated from the reaction mixture by centrifugation. 10 ml of pure water was added and only the gel part was isolated from this by centrifugation. This process was repeated four times whereby the unreacted antibody was removed. Then, the gel isolated in this way was treated with 2.1 ml of a blocking buffer (0.2 M Na phosphate, 1 M ethanolamine, pH 7.0) and 1.1 ml of the reducing agent solution added thereto, at 4 °C for 2 hours, whereby the the active groups in the unreacted gel were blocked. Finally, the resulting gel was washed with water, 20 mM Tris-HCl and 0.15 M NaCl (pH 8.0) using a centrifuge. This was filled into a small column tube, which was washed with 3 M sodium thiocyanate solution and 0.1 M glycine-HCl (pH 2.5) solution and then re-equilibrated with 20 mM Tris-HCl and 0.15 M NaCl (pH 8.0). The column thus equilibrated was stored.

The anti-TPO peptide antibody gel in the anti-HT1-2 antibody column had a binding efficiency of up to 94.2% for each IgG fraction. The amount of the IgG fraction linked to the gel per unit volume of the gel was 1.9 mg/ml gel. In the same way as above, a gel was formed in which 21.8 mg, per ml of the gel of IgG without TPO as antigen had been linked. This was used as a pre-column (referred to below as a pre-antibody column) to remove non-specifically bound molecules from the TPO antibody columns, as mentioned in (2). (2) Another example of preparing an anti-HT1-2 antibody column is illustrated below.

A standard sample of recombinant human TPO which has been partially purified from the supernatant of the culture of the CHO cells into which a gene encoding the amino acid sequence of any of SEQ ID NOS: 119-124 had been introduced and expressed was added to the pre-antibody column (containing 1 ml gel) prepared in (1), and 10 times the volume of the gel of 20 mM Tris-HCl (pH 8.0) and 0.15 M NaCl was passed through the column, whereby the fraction passing through the column was collected. Then, the fraction adsorbed to the pre-antibody column was eluted with 10 times the volume of the gel into an acidic eluent (0.1 M glycine-HCl, pH 2.5). Then the fraction that had passed through the preantibody column was added to the anti-HT1-2 antibody column (containing 2 ml of gel) prepared in (1) and the column was washed with 10 times the volume of the gel consisting of 20 mM Tris-HCl (pH 8.0) and 0.15 M NaCl, while the fraction passing through the anti-HT1-2 antibody column is collected. Then the fraction adsorbed to the anti-HT1-2 antibody column was eluted with 8 times the volume of the gel of an acidic eluent (0.1 M glycine-HCl, pH 2.5). These fractions were analyzed by SDS-PAGE which verified that TPO was not adsorbed to the pre-antibody column, but was specifically bound to the anti-HT1-2 antibody column, and that almost all added TPO in the sample was bound to the latter column. From this it was confirmed that at least from 200 to 300 ug/ml gel TPO was bound to the gel in the latter column.

EXAMPLE 72

Effect of TPO on increasing the number of platelets

20 normal ICR strain male mice, 8 weeks old (blood was already collected from their orbital veins and platelet counts were measured using a microcell counter F-800 manufactured by Toa Iyo Denshi) were randomly divided into four groups. One of the four groups (a control iv group) was injected with 100 µA PBS intravenously once daily for 5 consecutive days. One of the other groups (TPO-iv group) was injected with 100 µl of a solution) formed by replacing the concentrated solution of TPO active fraction F2 to SP Sepharose Fast Flow obtained in Example 56 with PSB using a NAP-25 column (Pharmacia Biotech; Catalog No. 17-0852-02) followed by dilution with PBS; containing relative activity, 211,900 according to the M-07e assay) intravenously once daily for 5 consecutive days. Another group (control-sc group) was injected with 100 microliters of PBS subcutaneously once daily for 5 consecutive days while the remaining group (TPO-sc group) was injected with 100 microliters once

daily for 5 consecutive days in the same way as in the TPO-iv group subcutaneously. After 6, 8, 10, 13 and 15 days from the initial administration, blood was collected from the orbital vein of each of the mice and the number of platelets was counted using a microcell counter (F-890; manufactured by Toa Iyo Denshi). Changes in the number of platelets are indicated in fig 24.1

in the control-iv group, the platelet count was increased by 44% compared to before administration even on day 6 (day 0 is the day of administration) where the platelet count was highest and, in the control-sc group, a 47% increase was noted even on day 8 where the numbers were highest . In the TPO-iv group, a 177% increase was recorded on day 6 compared to the numbers before administration and, in the TPO-sc group, an increase of 347% was noted already on day 6, and on day 8 the highest increase of 493% achieved.

Based on the above results, it has been confirmed that human TPO increases the number of platelets regardless of the difference in type and also of the route of administration.

EXAMPLE 73

Effect of TPO in increasing the number of platelets

16 normal male C3H/HeJ strain mice, 11 weeks old (blood had already been drawn from their orbital veins and platelet counts were measured using a microcell counter type F-800 manufactured by Toa Iyo Denshi) were divided into four random groups. One of four groups (a control group) was injected with 100 microliters of PBS subcutaneously once a day for 5 consecutive days. Another group (TPO-1 group) was injected subcutaneously with 100 microliters of a solution (prepared by substituting the TPO active fraction Sepha cryl S-200HR obtained in Example 56 with PBS using NAP-25 followed by dilution with PBS; containing relative activity, 83,000 according to M-07e

analysis) once daily for 5 consecutive days. Another group (the TPO-2 group) remained

i injected with 100 microliters of a solution (formed by diluting the TPO active fraction used in the TPO-1 group with PBS to the extent of 2 times; containing relative activity, 41,500 according to M-07e analysis) subcutaneously once a day for 5 following days. The last group (TPO-3 group) was injected with 100 microliters of a

solution (formed by dilution of the TPO active fraction used in TPO-2

> the group to a degree of 2 times; containing relative activity, 20.750 according to the M-07e assay) subcutaneously once daily for 5 consecutive days.

Blood was collected from the orbital veins of mice on days 6, 8, 10 and 12 after injection and counted

Platelets were counted using a microcell counter (type F-800; manufactured by

in Toa Iyo Denshi).

The changes in the number of platelets are given in fig 25.1 the control group there were almost no changes in the number of platelets while in groups injected with TPO there were

increases in the number of platelets in all groups to the extent that the numbers were higher on day 8 » after injection. Differences in dose-response were also noted between the groups. In TPO-1

group accordingly a 200% increase was already recorded on day 6 and the increase increased to approximately 270% on day 8 and thereafter a reduction was noted, even on day 12 approximately a 65% increase was still noticeable as there was still a significant difference from

the control group (p < 0.01; according to the Dunnet multiple comparison test). In the TPO-2 group

in the same increase was detected and, despite the increasing effect being less than the TPO-1 group, increases of approximately 140% and approximately 160% were noted on days 6 and 8. On day 12, the number was reduced to almost the same level as the control group. The increasing effect of the TPO-3 group was weaker than the TPO-2 group and on day 8

were the numbers the highest, approximately a 110% increase was indicated and it was again significant

in difference (p < 0.01) from the control group.

From the above results, it has been confirmed that human TPO increases platelet counts independent of mouse strain and that its effect has a dose-responsive nature.

EXAMPLE 74

An inhibitory effect of TPO on thrombocytopenia caused by anti-cancer drug administration

30 ICR male mice, 8 weeks old, were injected with 200 mg/kg 5-FU intravenously and randomly divided into two groups (each group includes 15 mice). From the next day (day 1), one of the groups (the control group) was injected with 100 microliters of PBS subcutaneously once a day for 5 consecutive days, while another group (TPO group) was injected with 100 microliters of the TPO active fraction used in Example 72 (containing relative activity, 211,900 according to the M-07e assay) subcutaneously once daily for five consecutive days. On days 4, 6 and 8, every 5th mouse was randomly selected from each group and then blood was collected from the orbital veins and the number of platelets was counted using a microcell counter (type E-2500; manufactured by Toa Iyo Denshi). Changes in the number of platelets are given in fig 26.1 the control group, the number of platelets after administration of 5-FU was reduced as

days passed and on day 6 the value was the lowest (about 28% of the value for administration of 5-FU), but on day 8 an about 1.3-fold increase was recorded in response to the reduction. Even in the case of a TPO group, the number of platelets after administration of 5-FU decreased as the days went by and on day 6 the value was the lowest (about 52% of the value before administration of 5-FU). The difference between the number of platelets on days 4 and 6 was very small, and on day 6 a significantly (p < 0.05; according to a "Dunnet multiple comparison test") high value appeared compared to the control group. On day 8 there was an increase of approximately 200% of the value before 5-FU administration.

From the above result, it has been confirmed that human TPO inhibits the reduction of the platelet count caused by the anti-cancer agent.

EXAMPLE 75

A therapeutic action of TPO for thrymbocytopenin

Nimustine hydrochloride (ACNU) (50 mg/kg) was injected intravenously into each of 36 ICR male mice that were 7 weeks old and then the mice were randomly divided into two groups (each group includes 18 mice). From the next day (day 1), one of the groups (the control group) was injected with 200 microliters of PBS subcutaneously twice a day for consecutive days, while another group (TPO group) was injected with 200 microliters of the TPO active fraction used in Example 73 (without dilution with PBS) to which 0.04% Tween 80 was added (containing relative activity, 380,000 according to the M-07e assay) subcutaneously twice daily on the following days.

After injection, the mice in each group were randomly divided into three groups and from one group blood was collected on days 5 and 12 and from another group on days 8 and 14 and from the remaining group on day 10. Blood was collected from the orbital veins and the number platelets were counted using a microcell counter (type F-800; manufactured by Toa Iyo Denshi). Changes in the number of platelets are given in fig 27.1 the control group the number of platelets after administration of ACNU decreased as the days went by and on days 8-10 the value was the lowest (about 29% of the value before administration of ACNU) and the achievement was only about 49% and about 74% on days 12 and 14 respectively. In contrast, in the TPO group a reduction was achieved to approximately 38% of the value before ACNU administration indicated on day 5 and then the situation changed to an increase. On day 8 the value became approximately 63% of the value before ACNU administration and on day 10 and thereafter the platelet count was greater than before ACNU administration. On days 12 and 14, increases of up to about 300% and about 400% were indicated.

As stated above, on days 8-10 where an increase was recorded in the control group, an increase was observed in the TPO group and on day 10 the platelet count was higher than those before the administration of ACNU. Accordingly, it has been confirmed that human TPO promotes healing from thrombocytopenia caused by the anti-cancer agent.

When the results according to example 74 are also taken into account, it is assumed that human TPO inhibits the reduction in the number of blood platelets and promotes healing from thrombocytopenia regardless of the type of anti-cancer agent.

EXAMPLE 76

A therapeutic effect of TPO on thrombocytopenia after BMT

48 male mice of the C3H/HeN strain that were 7 weeks old were irradiated with 10 Gy of radioactive irradiation on the whole body and immediately afterwards they were injected intravenously with 1x10^ bone marrow cells from mice of the same strain. The mice were then randomly divided into two groups and from the next day (day 1) one of the groups (the control group) was injected with PBS while another group (the TPO group) was injected with the TPO active fraction used in example 73 (containing relative activity 44,000 according to the M-07e assay) with a dose of 100 microliters each subcutaneously once daily for 20 consecutive days.

After initiation of injection, every 6 mice were randomly selected on days 5, 10, 14, and 21, blood was collected from their orbital veins, and the number of platelets was counted using a microcell counter (type E-2500; manufactured by Toa Iyo Denshi).

Changes in the number of platelets are indicated in fig 28.1 in all groups, the number of platelets decreased after the days passed after the application of BMT and on day 10 the number was the lowest (approximately 3% of the number before BMT application). After this, a gradual improvement was noted and on day 4 the number was approximately 11% and approximately 13% in the control group and in the TPO group. On day 21, a 37% cure was achieved in the control group, while in the TPO group the cure was approximately 65%, while a significant healing-promoting effect was observed. From the above results, it has been confirmed that human TPO formed as in Example 56 promotes healing of the platelet count after application of BMT.

EXAMPLE 77

A therapeutic effect of TPO for thrombocytopenia after irradiation with radioactive rays

36 male mice from the ICR strain aged 8 weeks were irradiated with X-rays of 5 Gy on the whole body and randomly divided into two groups. From the next day (day 1) one was off

the groups (control group) injected with PBS while another group (TPO group) was injected with TPO active fraction used in Example 73 (containing relative activity, 1,440,000 according to M-07e analysis) once daily for 3 consecutive days subcutaneously and from the next day, relative activity 360,000 according to the M-07e assay was injected subcutaneously once daily for 7 consecutive days. After the injection had begun, everyone in the group was randomly divided into 3 groups, a group where blood had been drawn on days 4, 11 and 21; another group on days 7 and 13 and the remaining group on days 9 and 15. Blood collection was performed from the orbital veins and the number of platelets was counted by a microcell counter (type F-800; manufactured by Toa Iyo Denshi). Changes in the number of platelets are indicated in Fig. 29.1 the control group, the number of platelets consequently decreased as the days passed after irradiation with X-rays, and, on day 9, the number was the lowest (about 25% of the number before irradiation with X-rays). On days 11 and 13 the numbers achieved were 38% and about 67% and on day 15 they returned to the values before irradiation with X-rays. In the TPO group, the number was reduced to approximately 24% of what it was before irradiation with X-rays on day 7. After this, an increase

observed on day 9 and the number improved to approximately 82%. On day 11 and thereafter, the platelet count was more than what appeared before irradiation with X-rays and until day 21 this tendency was maintained. As mentioned above on day 9 where the number was reduced in the control group, the number of platelets in the TPO group changed to an increase and on day 11 the number of platelets was higher than before irradiation with X-rays and remained high thereafter. In contrast, in the control group it took 15 days for the number of platelets to return to the same level as before irradiation with

x-rays. From these results, it was confirmed that the human TPO produced in Example 56 shortens the time required to recover from the reduction in the number of platelets after X-ray irradiation and exhibits a therapeutic effect for X-ray thrombocytopenia.

EXAMPLE 78

The effect of hTP0163 on increasing the number of platelets

15 healthy male mice of the C3H/HeN strain from which blood had already been collected from their orbital veins and subjected to platelet counts were measured using a microcell counter (type F-800; manufactured by Toa Iyo Denshi) were randomly divided into 4 groups - a control and groups A, B and C. The first group (control group) was injected subcutaneously with PBS containing 0.1% serum of mice once daily for 5 consecutive days. Groups A, B and C were injected with TPO active fraction from SP Sepharose Fast Flow obtained in Example 65 followed by dilution with PBS containing 0.1% mouse serum at doses of about 40,000,000, about 8,000,000 and about 1,600,000 kg /body weight/day with regard to relative activity to hTPO 163 according to the M-07e assay for 5 consecutive days subcutaneously.

On days 6, 8, 10 and 12 after injection, blood was collected from the orbital vein of each mouse and the number of platelets was counted using a microcell counter (type F-800; manufactured by Toa Iyo Denshi). The changes in the number of platelets are given in fig 30.

In the control group there were almost no changes in the number of platelets, while in the group given TPO increases in the number of platelets appeared in all cases with the highest values on day 8 after injection. A dose-response relationship was noted among each of the groups receiving TPO. Accordingly, in group A increases of approximately 88% and approximately 100% were noted on days 6 and 8 respectively, even on day 10 an increase of approximately 97% was maintained with all data showing a significant difference from the control group (p < 0.01 according to "Dunnet Multiple Comparison Test"). Similar increases also appeared in group B and despite the degree of increase being lower than that of group A, the increase of approximately 65% and approximately 84% was recorded on days 6 and 8 which shows a significant difference from the control group (p < 0.01 or 0.05 according to a "Dunnet Multiple Comparison Test"). The increasing effect in group C was lower than that in group B. It showed about 31% increase on day 8 when the highest numbers were recorded. From the above results, it has been confirmed that hTPO 163 produced in Example 65 increases the number of platelets and that the effect shows a dose-response relationship.

EXAMPLE 79

A therapeutic effect of hTP0163 for thrombocytopenia

100 male mice from the C3H/HeN strain at the age of 8 weeks were injected intravenously with 50 mg/kg nimustine hydrochloride (ACNU) and then randomly divided into four groups, a control group and groups A, B and C, each comprising 25 mice. The next day (day 1), the control group was injected subcutaneously with 5 ml/kg PBS once daily for consecutive days, while groups A, B and C were injected subcutaneously with TPO active fraction from SP Sepharose Fast Flow obtained in Example 65 diluted with PBS with a dose of about 16,000,000, about 48,000,000 and about 160,000,000/kg body weight/day, with respect to relative activity to hTPO 163 according to the M-07e assay once daily for consecutive days.

After initiation of injection, mice in each group were randomly divided into 5 groups and blood was collected on days 5, 8, 10, 12 and 14. Blood collection was performed from the orbital veins and the number of platelets was counted using a microcell counter (type E -2500; crafted by Toa Iyo Denshi). Changes in the number of platelets are shown in fig 31.1 the control group, the number of platelets decreased as the days went by after administration of ACNU, being the lowest values (approximately 15-16% of that before ACNU administration) on days 8-10 and the degree of healing was as little as approx. 28% and about 51% on days 12 and 14.1 group A, on the other hand, the number was reduced to about 25% compared to those before ACNU administration on day 8. After this, on the other hand, there was an increase and achievement of about 34% and about 89% on days 10 and 12. On day 14, the number was greater than before the ACNU administration.

As mentioned above, the lowest platelet counts were recorded on day 8 in all groups despite the fact that in groups given human TPO the reduction was slower than the control group which gives rise to the lowest number of platelets. In the control group, there was almost no increase in the number of platelets even on day 10, while in the groups when given human TPO, the increase in the number of platelets was recorded in all groups, although the degrees varied. In a group given 50 micrograms/kg, achievement was recorded on day 12 which already gave higher numbers than those before ACNU administration. When compared to the fact that in the control group the achievement was as low as about 51% compared to the condition before ACNU administration even on day 14 it is now clear that hTPO 163 produced as in Example 65 promotes healing from reduction in platelet count. Accordingly, hTPO 163 has been confirmed to have a therapeutic effect on thrombocytopenia caused by anti-cancer agents.

Examples of pharmaceutical preparations are given below.

EXAMPLE 80

Human TPO fraction obtained in Example 56 was concentrated, subjected to aseptic processing and frozen at -20°C to provide a frozen product which is used as an injectable preparation.

EXAMPLE 81

The human TPO fraction obtained in Example 56 was concentrated, 5 ml was filled into a 10 ml container using aseptic processing, lyophilized at -20 °C and the bottle was put on a rubber stopper to provide a lyophilized compound which can be used as an injectable preparation.

EXAMPLE 82

The human TPO fraction obtained in Example 56 was concentrated, subjected to aseptic filtration and filled into a 10 ml container to provide a product which can be used as an injectable preparation.

EXAMPLE 83

The hTPO 163 fraction obtained in Example 65 was concentrated, subjected to aseptic processing and lyophilized at -20°C to provide a frozen product which can be used as an injectable preparation.

EXAMPLE 84

The hTPO 163 fraction obtained in Example 65 was concentrated, 5 ml was filled into a 10 ml container using aseptic processing, lyophilized at -20 °C and the bottle was sealed with a rubber stopper to provide a lyophilized product which can be used as an injectable preparation.

EXAMPLE 85

The hTPO 163 fraction obtained in Example 65 was concentrated, subjected to aseptic filtration

and filled into a 10 ml container to provide a product which can be used as an injectable preparation.

EXAMPLE 86

The human TPO fraction obtained in Example 57 was concentrated, subjected to a

in aseptic processing and frozen at -20°C to provide a frozen product that can be used as an injectable preparation.

EXAMPLE 87

i The human TPO fraction obtained in Example 57 was concentrated, 5 ml was filled into a 10 ml container using aseptic method, lyophilized at -20 °C and the bottle was sealed with a rubber stopper to provide a lyophilized product which is used as an injectable preparation.

EXAMPLE 88

The human TPO fraction obtained in Example 57 was concentrated, subjected to aseptic filtration, filled into a 10 ml container to provide a product that can be used

as an injectable preparation.

in

EXAMPLE 89

Aplastic rabbit plasma

Heparinized aplastic rabbit plasma ("APK9") or normal rabbit plasma ("NK9") was produced as described [Mazur, E. and South, K. Exp. Hematol. 13:1164-1172 (1985);

Arriga, M., South K., Cohen J.L. and Mazur, E.M. Blood 69: 486-492 (1987); Mazur, E., Basilico, D., Newton, J.L., Cohen, J.L. Charland, C, Sohl, P.A. and Narendran, A. Blood

76: 1771-1782 (1990)] with the exception that 450 rad of total body irradiation was applied in the receivers.

EXAMPLE 90

Human megakaryocyte analysis

Standard methods for many of the procedures described in the following examples, or suitable alternative procedures, are provided in recognized molecular biology manuals such as Sambook et al (Eds.) Molecular Cloning, Second Edition, ColdSpring Harbor Laboratory Press (1987) and in Ausbel , et al, (Eds.), Current Protocols in Molecular Biology, Green associates/Wiley Interscience, New York (1990).

APK9 and fractionated APK9 were analyzed for the ability to stimulate development of human megakaryocytes from CD34<+> progenitor cells. CD34-selected cells were obtained from peripheral blood cells as described (Hokom, M.H., Choi, E., Nichol, J.L., Hornkohl, A., Arakawa, T., and Hunt, P. Molecular Biology of Haematopoiesis 3:15-31, 1994) and were incubated in the following culture medium: Iscove's modified Dulbecco's medium (IMDM; GIBCO, Grand Islan, NY) supplemented with 1% glutamine Pen-strep (Irvine Scientific, Santa Ana, CA) and 10% heparinized, platelet-poor, human AB plasma. Also included were 2-mercaptoethanol (10-<4> M), pyruvic acid (110 ug/ml), cholesterol (7.8 ug/ml), adenosineguanine, cytidine, uridine, thymidine, 2-deoxycytosine, 2-deoxyadenosine, 2 -deoxyganosine (10 ng/ml each, Sigma); human recombinant insulin (10 ng/ml), human transferrin (300 ng/ml), soybean lipids (1%, Boehringer Mannheim, Indianapolis, IN); human recombinant basic fibroblast growth factor (2 ng/ml), Genzyme, Cambridge, MA); human recombinant epidermal growth factor (15 ng/ml), platelet-derived growth factor (10 ng/ml, Amgen, INc, Thousand Oaks, CA). CD34-selected cells were seeded in 2 x 10<5> ml culture medium, 15 ni total volume, in wells of Terasaki-type microtiter dishes (Vanguard Inc, Neptune, NJ). The cells were incubated at 37 °C for 8 days in humidity boxes in 5% CO2 in air, fixed directly to culture wells with 1% glutaraldehyde and incubated with a monoclonal antibody cocktail (anti-GPIb, anti-GPITB, (Biodesign) and anti-GPIb ( Dako, Carpinteria, CA). The immunoreaction was developed with a streptavidin-beta-galactosidase detection system (HistoMark, Kirgegaard and Perry). Megakaryocytes identified by a blue stain were counted with an inverted phase microscope at 100 X magnification. Results were presented as mean number megakaryocytes per well ± standard deviation of the mean (SEM). In some cases, data were presented in terms of "megakaryocyte units/mr<1> where the degree by which a given sample induced megakaryocyte development was normalized to the positive APK9 control for the experiment. One unit is defined as the amount of material resulting in the same number of megakaryocytes as 1 ni APK9 standard The activity was accepted to be caused by MPL ligand if it cow nne be blocked with 5-10 ng/ml MPL-X (soluble Mpl receptor).

APK9 has been demonstrated to contain factors that stimulate the development of human megakaryocytes in this system. CD34-selected cells incubated with 10% NK9 for 8 days show a negligible number of blue-stained megakaryocytes, while CD34-selected cells incubated with 10% APK9 for 8 days show a very large number of blue-stained megakaryocytes.

Figure 32 shows that increasing concentrations of Mpl-X added to the human megakaryocyte culture system increasingly block megakaryocyte development. At concentrations of Mpl-X higher than 5 ug/ml, the inhibition is complete. In this experiment, CD34-selected cells were stimulated with 5% APK9. This demonstrates that an activity that reacts with Mpl-X (presumptive Mpl ligand) is required for the development of human megakaryocytes and indicates that the MPI ligand is present in APK9 itself.

It has further been demonstrated herein that the Mpl ligand activity required for the development of human megakaryocytes is present in APK9. APK9 (135 ml) was diluted 6-fold in Iscove's medium and applied to an Mpl-X affinity column. Unbound material (flow-through was collected and concentrated to original volume before analysis). The bound material was eluted in 10 ml of 1 M NaCl and 20% of the mixture was diafiltered and concentrated 4 times for analysis. CD34-selected cells incubated in medium alone did not develop into megakaryocytes. Cells incubated in 5% APK9 (same mixture as applied to the column) resulted in 48 ± 8 megakaryocytes per well. Cells incubated in 10% of the unbound material did not develop into megakaryocytes. Cells incubated in 10% of the elution mixture were developed to 120 ± 44 megakaryocytes per well. Both the activities of the column load and the elution mixture were substantially completely inhibited by 5 µg/ml Mpl-X in the assay.

EXAMPLE 91

Transfection of murine or human Mpl receptor into a murine cell line

A. Murine Mpl receptor

Full-length murine Mpl receptor cDNA was subcloned into an expression vector containing a transcriptional promoter derived from the Moloney murine sarcoma virus LTR. 5 µg of this construct and 1 µg of selectable marker plasmid pWLNeo (Stratagene) were co-electrophoresed into an IL-3 dependent murine cell line (32D, cline 23; Greenberger et al, PNAS 80: 2931-2936 (1983)). The cells were cultured for 5 days to recover from the procedure and then incubated in selection medium containing 800 ng/ml Geneticin (G418, Sigma) and 1 ng/ml murine IL-3. Surviving cells were then split into mixtures of 2 x 10<5> cells and cultured until a population emerged that could be further analyzed. Six populations were tested for surface expression of Mpl receptor by FACS analysis using a polyclonal rabbit antipeptide serum. A population was selected for FACS sorting by administration of the same antipeptide serum as before. Single-cell clones of the parent cell line were selected by growth in 10% APK9 and Geneticin. After selection in APK9 for 35 days, cells were maintained in 1 ng/ml murine IL-3. One of the subclones, 1 A6.1 was used for this work.

B. Human Mpl receptor

Full-length human Mpl receptor sequence (Vigon, I. et al, PNAS 89: 5640-5644 (1992)) was subcloned into an expression vector containing the transcription promoter of Maloney murine sarcoma virus (same vector as with murine receptor). 6 µg of this construct and 6 ng of an amphotrophic retroviral packaging construct (Landau, N.R., Littman, D.R., J. Virology 66: 5110-5113 (1992)) were transfected into 3 x 10 106 293 cells using a CaPC* 4 mammalian transfection kit, Stratagene. The same cells were transfected again after 2 days and again after 4 days. The day after the last transfection, the 293 cells were co-cultured with an IL-3 dependent murine cell line (32D, clone 23; Greenberger et al, PNAS 80: 2931-2936 (1983)). After 24 hours, the 32D cells were collected and separated in a BSA gradient. (Path-o-cyte; Mills Inc.). The cells were grown in a 1 ng/ml murine JL-3 and then selected for growth in 20% APK9. The cells were sorted for cell surface expression of receptor by FACS using a polyclonal rabbit antipeptide serum. These cells were then used in the analyses.

EXAMPLE 92

1 A6.1 assay for Mpl ligand

1A6.1 cells were washed free of culture IL-3 and replated (1000 cells/15 ni total col/well) in Terasaki-type microtitre dishes in alpha MEM (Gibco) supplemented with 10% fetal calf serum (FCS), Geneticin (800 ng /ml) and 1% pen/strep (Gibco) in 1:1 serial dilutions of the test samples. After 48 hours, the number of viable cells per well was determined microscopically. An activity unit was defined as the amount of activity that resulted in 200 viable cells per well. The activity was defined as caused by Mpl ligand if it could be completely blocked by including 5-10 ng/ml Mpl-X in the assay. Mpl ligand activity in APK9 averaged 4400 ± 539 units/ml aplastic plasma. If not stated otherwise, units of Mpl ligand activity are defined in the 1A6.1 analysis.

The analyzes with cells transfected with the human Mpl receptor gene were performed essentially in the same way as with the 1A6.1 cells.

EXAMPLE 93

Demonstration that the Mpl ligand is present in aplastic plasma of sera from mouse, dog, pig and human sources

The Mpl ligand is present in aplastic plasma or sera from murine, rabbit, pig and human sources (Table 9). Plasma was collected from BDF1 mice pre-irradiated and 12 days post-irradiation (500 rad). Plasma was tested in the 1A6.1 assay where 2000 units/ml activity was demonstrated which was substantially completely inhibitable with Mpl-X (10 µg/ml). Irradiated mouse plasma was also positive in human megakaryocyte analysis where it showed an activity of 1833 units/ml. Plasma was collected from dogs pre-irradiated and 10 days post-irradiation (450 rad). Plasma was tested in both the 1 A6.1 assay and human megakaryocyte assays. The activity was detected and completely inhibited by Mpl-X (10 µg/ml) in both assays. Plasma was collected from pigs pre-irradiated and 10 days after irradiation (650 rad). Plasma was tested in both the 1 A6.1 assay and human megakaryocyte assays. Both assays showed Mpl ligand activity (inhibitable by 10 ng/ml Mpl-X) comparable to that found in aplastic rabbit plasma. Sera from aplastic humans were obtained. This material was collected from bone marrow transplant patients. Sera from 6 patients were analyzed in the 1A6.1 assay where it showed an activity of 903 units/ml, 88% was caused by Mpl ligand (inhibitable with 10 ng/ml Mpl-X). Sera from 14 aplastic patients have also been tested in human megakaryocyte analysis. As a group, substantial activity was obtained, 941 meg units/ml, which was completely inhibitable with 10 ng/ml Mpl-X. Murine IL-3 data are included to demonstrate the specificity of the 1 A6.1 assay. Although this recombinant cytokine induces the growth of the cell line, it is not blocked by 10 ng/ml Mpl-X.

EXAMPLE 94

Preparation of human TPO derivatives expressed in the E.coli system

In an attempt to improve the biological activity, stability and solubility, several TPO derivatives were constructed and expressed in E.coli. The derivatives were formed using the amino acid sequence as set forth in SEQ ID NO: 11, including [Met-<2>, Lys"<1>, Ala<1>, Val3, Arg<12?>]TPO(1-163) where an arginine replaces a leucine at amino acid position 129 (also referred to as L129R), [Met"<2>, Lys"<1>, Ala<1>, Val<3>, Arg<133>]TPO(l-163) wherein an arginine replaces a histidine at amino acid position 133 (also referred to as H133R), [Met"2, Lys-1, Ala<1>, Val<3>, Arg<143>]TPO(l-163) wherein an arginine replaces a methionine at amino acid position 143 (also referred to as M143R) [Met"<2>, Lys"<1>, Ala<1>, Val<3>, Arg<82>]TPO(1-163) where a leucine replaces a glycine at amino acid position 82 (also referred to as G82L) [Met"<2>, Lys"<1>, Ala1, Val3, Arg<146>]TPO(l-163) where a leucine replaces a glycine at amino acid position 146 (also referred to as G146L) [Met-<2>, Lys" 1, Ala1, Val<3>, Arg<148>]TPO(l-163) where a proline replaces a serine at amino acid position 148 (also referred to as S148P), [Met"<2>, Lys"<1>, Ala<1>, Val<3>, Arg59]TPO(l-163) where an arginine replaces a lysine at amino acid position 59 (also referred to as K59R), [Met"<2>, Lys"<1>, Ala<1>, Val<3>, Arg< 115>]TPO(1-163) where an arginine replaces a glutamine at amino acid position 115 (also referred to as Q1 15R).

When forming the TPO derivatives, h6T (1-163), described in example 42, was used as a template, and the oligonucleotides that have the following sequences were synthesized: L129R: 5'-ATCTTCCGTTCTTTCCAGCACCT-3' (for the preparation of the Arg-substitution derivative with Leu- 129 substituted with Arg in the sequence shown in SEQ ID NO: 11);

H133R: 5-TCTTTCCAGCGTCTGCTGCGT-3' (preparation of the Arg substitution derivative with His-133 substituted with Arg in the sequence shown in SEQ TD NO: 11);

M143R: 5'-CGTTTCCTGCGTCTGGTTGGC-3' (for preparation of the Arg substitution derivative with Met-143 substituted with Arg in the sequence shown in SEQ TD NO: 11);

G82L: 5'-æCCAGCTTCTGCCGACCTGCCT-3' (for the preparation of the Leu substitution derivative with Gly-82 substituted with Leu in the sequence shown in SEQ ID NO: 11);

G146L: S-ATGCTGGTTCTGGGTTCTACCCT-S' (for preparing the Leu substitution derivative with Gly-146 substituted with Leu in the sequence shown in SEQ ID NO: 11);

S148P: 5'-GTTGGCGGTCCGACCCTGTGCG-3' (for preparation of the Pro substitution derivative with Ser-148 substituted with Pro in the sequence shown in SEQ ID NO: 11);

K59R: 5'-GAAGAGACCCGCGCTCAGGACATCC-3' (for the preparation of the Arg substitution derivative with Lys-59 substituted with Arg in the sequence shown in SEQ ID NO: 11);

Q115R: 5,-CTGCCGCCACGTGGCCGTACCAC-3, (for the preparation of the Arg substitution derivative with Gln-115 substituted with Arg in the sequence shown in SEQ ID NO: 11);

Mutant plasmids were constructed using the Sculptor in vitro mutagenesis kit (Amersham). The XbaI-HindlJJ gene derived from pCFM536/h6T(1-163) described in Example 42 was introduced into the Bluescript II SK(-) phagemid vector (Stratagene, California, USA) after cleavage with XbaI and Hindin to obtain a plasmid pSKTPO which then was transformed into E.coli JM109. Single-stranded DNA for mutagenesis was generated from pSKTPO using helper M13K07 (Takara Shuzo, Japan). The mutations were introduced into the TPO gene according to the vendor's protocols using the oligonucleotides as indicated above. Sequence analyzes were performed with the DNA sequencing kit (Applied Biosystems, USA) according to the vendor's protocols and the mutations to TPO were confirmed.

The XbaI-Hindm fragments generated from mutated pSKTPO were introduced into pCFM536 cleaved with XbaI and Hindin. The plasmids pCFM536 containing mutated genes were transformed into E.coli nr 261 to obtain the transformant expressing the TPO derivative genes.

Cultivation of E. coli No. 261 transformed with mutated pCFM 536 was carried out according to Example 43 as previously described. The cell pellet of the transformant was collected and stored in

the freezer for at least a day. The frozen cell pellet (3 g) was suspended in 30 ml of 20 mM Tris buffer (pH 8.5) containing 10 mM EDTA, 10 mM DTT and 1 mM PMSF. The suspension was kept on ice and sonicated 5 times at 1 minute (at least) intervals. Sonicated cells were collected by centrifugation at 15,000 rpm for 10 min.

The pellet was suspended in 30 ml of 10 mM Tris buffer (pH 8.7) containing 8 M

guanidine chloride, 5 mM EDTA, 1 mM PMSF and reduced by addition of 50 mg DTT. After stirring for 1-1.5 hours, the pH of the solution was adjusted to 5.0 using dilute HCl and was then cooled to 4°C before dilution.

The same solution was gradually diluted in 1.5 110 mM CAPS buffer (pH 10.5) containing 30% glycerol, 3 M urea, 3 mM cystamine and 1 mM L-cysteine overnight. After the sample was stirred for at least 2 days, the precipitate was removed by centrifugation at 8000 rpm for 45 min. The pH of the solution was adjusted to 6.8 with 6 M phosphoric acid and diluted twice. The solution was filtered with 2 sheets of No. 2 filter paper (ø-90 mm, Toyo filter paper, Japan) and then applied to the column (2.6 x 10 cm) CM-Sepharose Fast

Flow (Pharmacia). The column was washed with 400 ml of 10 mM phosphate buffer (pH 6.8)

in containing 15% glycerol and 1 M urea, and equilibrated with 10 mM phosphate buffer (pH

7.2) containing 15% glycerol. The protein was eluted from the column using a linear gradient system from 10 mM phosphate buffer (pH 7.2) containing 15% glycerol in the same buffer containing 0.5 M NaCl at a flow rate of 1.0 ml/min. Protein elution was recorded by absorbance at 280 nm and the fraction containing the TPO derivative was confirmed by SDS-PAGE and collected.

After the TPO fraction (approximately 30 ml) was concentrated to 2 ml using Centriprep 10 (Amicon), mutant TPO was further purified by reverse phase HPLC (Waters) with a column of uBondasphere C4 (3.9 x 150 mm). The protein elution was performed using a linear gradient from 0.05 TFA in H2O to 2-propanol containing 30% CH3CN and 0.02% TFA at a flow rate of 0.5 ml/min for 40 min. Under these conditions, the TPO derivatives were eluted after approximately 30 min.

For biological analysis, the TPO sample that has thus been purified was dialyzed twice against 1110 mM phosphate buffer for 2 days. After concentration with Centriprep 10 (Amicon) to approximately 0.1 mg/ml, the biological activity of the derivatives was assessed in the M-07e assay system as previously described. It was discovered that all the mutants have almost identical activity to h6T(1-163), the reference sample of TPO (see Figures 33 and 34).

Deposit list

Escherichia coli (pHTl-231/DH5): FERM BP-4564 and CCTCC-M95001 Escherichia coli (pEF18S-A2-o/DH5): FERM BP-4565 and CCTCC-M95002 Escherichia coli (pHGTl/DH5): FERM BP-4616 and CCTCC-M95003 Escherichia coli (pHTFl/DH5): FERM BP-4617 and CCTCC-M95004 Mouse-Mouse hybridoma P55: FERM BP-4563 and CCTCC-C95001 CH028/1/1/3-C6: FERM BP-4088 and CCTCC -C95004 CH0163T-63-79-C1: FERM BP-4989 and CCTCC-C95005

Claims (39)

1. Thrombopoietin (TPO) polypeptide, characterized in that it has biological activity that specifically stimulates or increases platelet production and comprises the amino acid sequence 1 to 332 according to SEQ ID NO: 6 or a polypeptide encoded for by a polynucleotide that hybridizes under stringent conditions with a polynucleotide that codes for the amino acid sequence SEQ ID NO:6. 2. TPO polypeptide derivative, characterized in that it consists of the amino acid sequence 1 to 163 according to SEQ ID NO: 6. 3. TPO polypeptide derivative, characterized in that it consists of the amino acid sequence 1 to 232 according to SEQ ID NO: 6. 4. TPO polypeptide derivative, characterized in that it consists of the amino acid sequence 1 to 151 according to SEQ ID NO: 6. 5. TPO polypeptide derivative according to claim 1, 2 or 3, characterized in that from 1 to 6 amino terminal amino acids have been deleted. 6. TPO polypeptide derivative according to claim 2, characterized in that it is selected from the group consisting of [AHis<33>]TPO (1-163), [AArg<117>]TPO (1-163) and [AGly<116>]TPO ( 1-163). 7. TPO polypeptide derivative according to claim 2, characterized in that it is selected from the group consisting of [His<33>, Thr<33>', Pro<34>]TPO (1-163), [His<33>, Ala33', Pro34 ]TPO (1-163), [His33, Gly33', Pro3<4>, Ser3<8>]TPO (1-163), [Gly<116>, Asn116', Arg11<7>]TPO (1-163 ), [Gly11<6>, Ala11<6>', Arg^jTPO (1-163), [Gly<116>, Glyl \ 6\ Argl 17]Tpo (1-163) and [Ala1, Val<3> ]TPO (1-163). 8. TPO polypeptide derivative according to claim 1, characterized in that it is selected from the group consisting of [Thr<33>, Thr<333>, Ser334, Ile335, Gly336, Tyr337, Pro338, Tyr33?, Asp34(>, Val341, Pro342, Asp343, Tyr344, Ala345, Gly346, Val347, His34<8>, His34?His35<0>, His3<*!>, His3<52>, His<353>]TPO and [Asn25, Lys231, Thr333, Ser334, Ile335, Gly336 , Tyr337, Pro338, Tyr33?, Asp34**, Val341, Pro342, Asp343, Tyr344, Ala345, Gly346, Val347, His348, His34?, His350, His35*, His352, His3<53>]TPO. 9. TPO polypeptide derivative according to claim 1, characterized in that it is selected from the group consisting of [Asn<25>]TPO and [Thr<33>]TPO. 10. TPO polypeptide derivative according to claim 2, characterized in that it is selected from the group consisting of [Ala<1>, Val<3>, Arg<12?>]TPO (1-163), [Ala<1>, Val<3> , Arg133]TPO (1-163), [Ala<1>, Val<3>, Arg<143>]TPO (1-163), [Ala<1>, Val3, Leu82]TPO (1-163), [Ala<1>, Val<3>, Leu14<6>]TPO (1-163), [Ala1, Val<3>, Pro<148>]TPO (1-163), [Ala1, Val3, Arg5? ]TPO (1-163) and [Ala<1>, Val<3>, Arg115]TPO (1-163). 11. TPO polypeptide derivative according to claim 2, characterized in that it is selected from the group consisting of [Arg<12?>]TPO (1-163), [Arg133]TPO (1-163), [Arg143]TPO (1-163), [Leu82]TPO (1-163), [Leu146]TPO (1-163), [Pro148]TPO (1-163), [Arg5?]TPO (1-163) and [Arg11<5>]TPO (1 -163). 12. TPO polypeptide according to claim 1, characterized in that it is covalently bound to a polymer. 13. TPO polypeptide according to claim 12, characterized in that said polymer is polyethylene glycol. 14. TPO polypeptide according to any one of claims 1 to 13, characterized in that it further comprises the amino acid Met"2-Lys" V 15. TPO polypeptide according to any one of claims 1 to 13, characterized in that it further comprises Met" 1. 16. TPO polypeptide according to any one of claims 1 to 13, characterized in that it further comprises Gly<1>. 17. Polypeptide, characterized in that it consists of the mature sequence according to the amino acids in SEQ TD NO: 2,4 or 6. 18. DNA, characterized in that it codes for a TPO polypeptide according to any one of claims 1 to 11, 14, 15 and 17 inclusive. 19. DNA sequence, characterized in that it is for use to ensure expression of a TPO polypeptide having biological activity comprising specific stimulation or increased platelet production, which is selected from the group consisting of: (a) the DNA sequences shown in SEQ ID NOS: 194,195 or 196 or complementary strands thereof; and (b) DNA sequences that hybridize under stringent conditions to the DNA sequences as defined in (a) or fragments thereof; and (c) DNA sequences that can hybridize to the DNA sequences as defined in (a) and (b), excluding the degeneracy of the genetic code. 20. DNA sequence according to claim 19, characterized in that said sequence is cDNA. 21. DNA sequence according to claim 19, characterized in that said sequence is genomic DNA. 22. DNA sequence according to claim 19, characterized in that said sequence is manufactured DNA. 23. Method for producing a TPO polypeptide, characterized in that it comprises the following steps: (a) expressing a polypeptide which is encoded by a DNA according to any one of claims 18, 19, 20, 21 or 22 in a suitable host and (b) isolating said TPO polypeptide. 24. Method according to claim 23, characterized in that the TPO polypeptide that is expressed is a Mef^-Lys-1 polypeptide and includes further cleavage of Met'<2->Lys"<1> from said isolated TPO polypeptide. 25. DNA sequence according to claim 19, characterized in that said sequence further codes for a thrombin recognition peptide 5' for TPO polypeptide coding sequences and codes for glutathione S-transferase (GST) 5' to the thrombin recognition peptide. 26. DNA sequence according to claim 25, characterized in that it is for use to ensure expression of a TPO polypeptide consisting of the amino acid sequence 1 to 174 according to SEQ ID NO: 6. 27. Process for producing a TPO polypeptide with biological activity comprising specifically stimulating or increasing platelet production, characterized in that it includes the following steps: (a) expressing a polypeptide encoded by a DNA according to claim 25 or claim 26 in a suitable host, and (b) isolating said GST-(thrombin recognition peptide)-TPO polypeptide, (c) treating said isolated GST-(thrombin recognition peptide)-TPO polypeptide with thrombin and (d) isolating a Glyl .jpo polypeptide. 28. Method according to claim 27, characterized in that the Gly-^TPO polypeptide is [Gly^TPO (1-174). 29. Method according to claim 27 or claim 28, characterized in that the isolated Gly1-TPO polypeptide is a TPO derivative consisting of the amino acid sequence 1 to 174 according to SEQ ID NO: 6. 30. Protein product, characterized in that it is obtained according to the method in claims 23, 24, 27, 28 or 29. 31. Prokaryotic or eukaryotic host cell, characterized in that it is transformed or transfected with the DNA sequence according to any one of claims 18, 19, 20, 21, 22, 25 or 26 in a way that enables said host cell to express a polypeptide that has biological activity involving specific stimulation or increasing platelet production. 32. Pharmaceutical composition, characterized in that it comprises an effective amount of the polypeptide according to any one of claims 1 to 16 and 30 in combination with a pharmaceutically acceptable carrier. 33. Pharmaceutical composition according to claim 32, characterized in that it is for use in the treatment of platelet disorders. 34. Pharmaceutical composition according to claim 32, characterized in that it is for use in the treatment of thrombocytopenia. 35. Pharmaceutical composition according to claim 34, characterized in that it is for use in the treatment of thrombocytopenia induced by chemotherapy, radiotherapy or bone marrow transplantation. 36. Use of the polypeptide according to any one of claims 1 to 16 and 30 for the manufacture of a medicament for the treatment of a platelet disorder in a patient. 37. Use according to claim 36, wherein the platelet disorder is thrombocytopenia. 38. Use according to claim 37, where said thrombocytopenia is induced by chemotherapy, radiotherapy or bone marrow transplantation. 39. Antibody, characterized in that it is specifically immunoreactive with a polypeptide according to any one of claims 1 to 16 and 30 inclusive.