The Glycan Domain of Thrombopoietin (TPO) Acts intrans to Enhance Secretion of the Hormone and Other Cytokines
2002; Elsevier BV; Volume: 277; Issue: 38 Linguagem: Inglês
10.1074/jbc.m201297200
ISSN1083-351X
AutoresHannah M. Linden, Kenneth Kaushansky,
Tópico(s)Chronic Lymphocytic Leukemia Research
ResumoThrombopoietin (TPO), the primary regulator of platelet production, is composed of an amino-terminal 152 amino acids, sufficient for activity, and a carboxyl-terminal region rich in carbohydrates (183 residues) that enhances secretion of the molecule. Full-length TPO is secreted at levels 10–20-fold greater than truncated TPO. By introducing into mammalian cells a novel cDNA encoding the TPO secretory leader linked to its carboxyl-terminal domain (TPO glycan domain (TGD)), we tested whether TGD could function in trans to enhance secretion of TPO. The artificial TGD was secreted, inactive in proliferation assays, and did not inhibit TPO activity. However, when co-transfected with a cDNA encoding truncated TPO, TGD enhanced secretion 4-fold, measured by specific bioassay and immunoassay. TGD also enhanced secretion of granulocyte monocyte colony-stimulating factor and stem cell factor but did not affect the production of erythropoietin, interleukin-3, growth hormone, or of full-length TPO. To localize TGD function, we added an endoplasmic reticulum (ER) retention signal to TGD and, separately, deleted the secretory leader. Deletion of the secretory leader attenuated the secretory function of TGD, whereas addition of the ER retention signal did not alter its function. To investigate the physiologic role of TGD in folding and proteasomal protection, we tested full-length and truncated TPO in assays of protein refolding, and we examined protein stability in the presence of proteasome inhibitors. We found that truncated TGD re-folds readily and that proteasome-mediated degradation contributes to the poor secretion of truncated TPO. We conclude that TGD enhances secretion of TPO and can additionally function as an inter-molecular chaperone, in part because of its ability to prevent degradation of the hormone. The cellular location of TGD action is likely to be within the ER or earlier in the secretory pathway. Thrombopoietin (TPO), the primary regulator of platelet production, is composed of an amino-terminal 152 amino acids, sufficient for activity, and a carboxyl-terminal region rich in carbohydrates (183 residues) that enhances secretion of the molecule. Full-length TPO is secreted at levels 10–20-fold greater than truncated TPO. By introducing into mammalian cells a novel cDNA encoding the TPO secretory leader linked to its carboxyl-terminal domain (TPO glycan domain (TGD)), we tested whether TGD could function in trans to enhance secretion of TPO. The artificial TGD was secreted, inactive in proliferation assays, and did not inhibit TPO activity. However, when co-transfected with a cDNA encoding truncated TPO, TGD enhanced secretion 4-fold, measured by specific bioassay and immunoassay. TGD also enhanced secretion of granulocyte monocyte colony-stimulating factor and stem cell factor but did not affect the production of erythropoietin, interleukin-3, growth hormone, or of full-length TPO. To localize TGD function, we added an endoplasmic reticulum (ER) retention signal to TGD and, separately, deleted the secretory leader. Deletion of the secretory leader attenuated the secretory function of TGD, whereas addition of the ER retention signal did not alter its function. To investigate the physiologic role of TGD in folding and proteasomal protection, we tested full-length and truncated TPO in assays of protein refolding, and we examined protein stability in the presence of proteasome inhibitors. We found that truncated TGD re-folds readily and that proteasome-mediated degradation contributes to the poor secretion of truncated TPO. We conclude that TGD enhances secretion of TPO and can additionally function as an inter-molecular chaperone, in part because of its ability to prevent degradation of the hormone. The cellular location of TGD action is likely to be within the ER or earlier in the secretory pathway. Thrombopoietin (TPO) 1The abbreviations used are: TPO, thrombopoietin; RBD, receptor binding domain; TGD, TPO glycan domain; EPO, erythropoietin; GM-CSF, granulocyte monocyte colony-stimulating factor; SCF, stem cell factor or Kit ligand; IL-3, interleukin-3; MK, megakaryocyte; TGF, transforming growth factor; SDF-1, stromal cell-derived factor 1; ERAD, endoplasmic reticulum-associated degradation; BHK, baby hamster kidney; MTT, 3[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; ELISA, enzyme-linked immunosorbent assay; ER, endoplasmic reticulum; GdnHCl, guanidine hydrochloride; MAPK, mitogen-activated protein kinase; CAT, chloramphenicol acetyltransferase; SL, secretory leader; GH, growth hormone; hGH, human GH; IP, immunoprecipitation; RT, reverse transcriptase; vWF, von Willebrand factor; mTPO, murine TPO. is the principal cytokine that regulates megakaryocyte development and platelet production (1Kaushansky K. N. Engl. J. Med. 1998; 339: 746-754Crossref PubMed Scopus (325) Google Scholar). TPO acts at both early (2Zeigler F.C. de-Sauvage F. Widmer H.R. Keller G.A. Donahue C. Schreiber R.D. Malloy B. Hass P. Eaton D. Matthews W. Blood. 1994; 84: 4045-4052Crossref PubMed Google Scholar) and late stages of megakaryopoiesis (3Debili N. Wendling F. Katz A. Guichard J. Breton Gorius J. Hunt P. Vainchenker W. Blood. 1995; 86: 2516-2525Crossref PubMed Google Scholar, 4Kaushansky K. Blood. 1995; 86: 419-431Crossref PubMed Google Scholar), alone and in synergy with other cytokines (5Broudy V.C. Lin N.L. Kaushansky K. Blood. 1995; 85: 1719-1726Crossref PubMed Google Scholar). The hormone also acts in synergy with erythropoietin (EPO) to stimulate erythropoiesis (6Kaushansky K.K. Broudy V.C. Grossman A. Humes J. Lin N. Ren H.P. Bailey M.C. Papayannopoulou T. Forstrom J.W. Sprugel K.H. J. Clin. Invest. 1995; 96: 1683-1687Crossref PubMed Scopus (225) Google Scholar). Subsequent studies have revealed TPO to be both necessary and sufficient for full MK1 maturation (6Kaushansky K.K. Broudy V.C. Grossman A. Humes J. Lin N. Ren H.P. Bailey M.C. Papayannopoulou T. Forstrom J.W. Sprugel K.H. J. Clin. Invest. 1995; 96: 1683-1687Crossref PubMed Scopus (225) Google Scholar). As a therapeutic agent TPO has been shown to speed platelet recovery following myelosuppressive therapy in cancer patients receiving chemotherapy (7Vadhan-Raj S. Verschraegen C.F. Bueso-Ramos C. Broxmeyer H.E. Kudelka A.P. Freedman R.S. Edwards C.L. Gershenson D. Jones D. Ashby M. Kavanagh J.J. Ann. Intern. Med. 2000; 132: 364-368Crossref PubMed Scopus (151) Google Scholar,8Basser R.L. Underhill C. Davis I. Green M.D. Cebon J. Zalcberg J. MacMillan J. Cohen B. Marty J. Fox R.M. Begley C.G. J. Clin. Oncol. 2000; 18: 2852-2861Crossref PubMed Scopus (54) Google Scholar). In addition, the biological effects of TPO are not limited to the MK lineage; the growth of erythroid (burst forming unit-erythrocyte) and myeloid (colony forming unit-granulocyte macrophage) colony-forming cells is also expanded by TPO in vitro, and its use in normal and myelosuppressed mice and non-human primates leads to enhanced recovery of multiple hematopoietic lineages (9Kaushansky K. Lin N. Grossmann A. Humes J. Sprugel K. Broudy V.C. Exp. Hematol. 1996; 24: 265-269PubMed Google Scholar, 10Akahori H. Shibuya K. Obuchi M. Nishizawa Y. Tsuji A. Kabaya K. Kusaka M. Ohashi H. Tsumura H. Kato T. Miyazaki H. Br. J. Haematol. 1996; 94: 722-728Crossref PubMed Scopus (50) Google Scholar). Finally, TPO affects the survival and proliferation of primitive hematopoietic stem cells in vitro (11Sitnicka E. Lin N. Priestly G.V. Broudy V.C. Wolf N.S. Kaushansky K. Blood. 1996; 87: 4998-5005Crossref PubMed Google Scholar, 12Kobayashi M. Laver J.H. Kato T. Miyazaki H. Ogawa M. Blood. 1996; 88: 429-436Crossref PubMed Google Scholar) and in vivo (13Solar G.P. Kerr W.G. Zeigler F.C. Hess D. Donahue C. de Sauvage F.J. Eaton D.L. Blood. 1998; 92: 4-10Crossref PubMed Google Scholar). In this manner, clinical benefit may be derived from the administration of TPO to enhance peripheral blood stem cell collection (14Somlo G. Sniecinski I. ter Veer A. Longmate J. Knutson G. Vuk- Pavlovic S. Bhatia R. Chow W. Leong L. Morgan R. Margolin K. Raschko J. Shibata S. Tetef M. Yen Y. Forman S. Jones D. Ashby M. Fyfe G. Hellmann S. Doroshow J.H. Blood. 1999; 93: 2798-2806Crossref PubMed Google Scholar) and theex vivo expansion of primitive hematopoietic cells (15Piacibello W. Sanavio F. Garetto L. Severino A. Bergandi D. Ferrario J. Fagioli F. Berger M. Aglietta M. Blood. 1997; 89: 2644-2653Crossref PubMed Google Scholar). TPO may thus offer patients with primary and acquired hematological disorders a significant therapeutic benefit. Thrombopoietin is a 335-amino acid polypeptide produced in multiple organs, first cloned as a ligand for the orphan hematopoietic cytokine receptor proto-oncogene c-mpl. Like EPO, GH, and other members of the hematopoietic cytokine family, the amino-terminal region of TPO is predicted to fold into a left-handed four-helix bundle protein. The amino acids of this domain of TPO (residues 1–152) share greater homology with EPO than does any other pair of hematopoietic cytokines (22% identity, and an additional 24% sequence similarity (16Lok S. Kaushansky K. Holly R.D. Kuijper J.L. Lofton-Day C.E. Oort P.J. Grant F.J. Heipel M.D. Burkhead S.K. Kramer J.M. et al.Nature. 1994; 369: 565-568Crossref PubMed Scopus (1042) Google Scholar)). Our laboratory and others (17Bartley T.D. Bogenberger J. Hunt P., Li, Y.S., Lu, H.S. Martin F. Chang M.S. Samal B. Nichol J.L. Swift S. et al.Cell. 1994; 77: 1117-1124Abstract Full Text PDF PubMed Scopus (920) Google Scholar, 18Hunt P., Li, Y.S. Nichol J.L. Hokom M.M. Bogenberger J.M. Swift S.E. Skrine J.D. Hornkohl A.C., Lu, H. Clogston C. et al.Blood. 1995; 86: 540-547Crossref PubMed Google Scholar, 19Linden H.M. Kaushansky K. Biochemistry. 2000; 39: 3044-3051Crossref PubMed Scopus (35) Google Scholar) have demonstrated that the amino-terminal (or cytokine-like) domain of TPO is adequate for receptor binding, signaling, and supporting cellular proliferation. In addition to the four-helical bundle domain, and unlike all other known members of the hematopoietic cytokine family, the TPO gene encodes a carboxyl-terminal polypeptide extension (residues 153–335), a serine- and a proline-rich domain remarkable for abundant carbohydrate modification, and a lack of homology to other known proteins. Glycosylation of this carboxyl-terminal (or glycan) domain has been experimentally determined (residues 153–246) and predicted (residues 246–332) in human TPO (20Hoffman R.C. Andersen H. Walker K. Krakover J.D. Patel S. Stamm M.R. Osborn S.G. Biochemistry. 1996; 35: 14849-14861Crossref PubMed Scopus (52) Google Scholar). The glycosylation of TPO accounts for approximately one-half of its observed 70-kDa molecular mass. Not surprisingly, the inter-species (mouse-human) homology of TPO is greatest in the amino-terminal receptor-binding domain (93%); however, the carboxyl-terminal domain retains 74% homology, suggesting that it also serves an important physiologic function. In previous work our laboratory and others (19Linden H.M. Kaushansky K. Biochemistry. 2000; 39: 3044-3051Crossref PubMed Scopus (35) Google Scholar, 21Muto T. Feese M.D. Shimada Y. Kudou Y. Okamoto T. Ozawa T. Tahara T. Ohashi H. Ogami K. Kato T. Miyazaki H. Kuroki R. J. Biol. Chem. 2000; 275: 12090-12094Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 22Ahn H.K. Chung J.Y. Park S.K. Joo S.M. Kook P.S. Koh Y.W. Biochem. Mol. Biol. Int. 1999; 47: 729-733PubMed Google Scholar) have shown that the TPO glycan domain (TGD) functions to enhance secretion; in our hands deletion of the TGD reduced TPO secretion ∼20-fold. Interestingly, several bacterial proteases display a two-domain structure, a protease domain, and a proregion essential for secretion, shown to facilitate the folding of the parent compound prior to extracellular proteolytic cleavage (reviewed in Refs. 23Eder J. Fersht A.R. Mol. Microbiol. 1995; 16: 609-614Crossref PubMed Scopus (165) Google Scholar and 24Baker D. Shiau A.K. Agard D.A. Curr. Opin. Cell. Biol. 1993; 5: 966-970Crossref PubMed Scopus (153) Google Scholar). Furthermore, several of these proregions have been studied and shown to function both in cis (covalently linked to their corresponding enzyme) and in trans (when co-expressed from a different gene). Co-transfection of cDNAs encoding a normal proregion with a cDNA encoding a protein either lacking a proregion or with a mutated proregion can rescue the protein from aggregation or destruction and hence facilitate protein secretion (25Silen J.L. Agard D.A. Nature. 1989; 341: 462-464Crossref PubMed Scopus (195) Google Scholar, 26Fabre E. Tharaud C. Gaillardin C. J. Biol. Chem. 1992; 267: 15049-15055Abstract Full Text PDF PubMed Google Scholar, 27Fukuda R. Horiuchi H. Ohta A. Takagi M. J. Biol. Chem. 1994; 269: 9556-9561Abstract Full Text PDF PubMed Google Scholar, 28van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar, 29Beer H.D. Wohlfahrt G. Schmid R.D. McCarthy J.E. Biochem. J. 1996; 319: 351-359Crossref PubMed Scopus (69) Google Scholar, 30Baier K. Nicklisch S. Maldener I. Lockau W. Eur. J. Biochem. 1996; 241: 750-755Crossref PubMed Scopus (17) Google Scholar). However, in some instances, the proregion is unable to rescue the parent protein to enhance secretion in trans (31Valverde V. Delmas P. Kaghad M. Loison G. Jara P. J. Biol. Chem. 1995; 270: 15821-15826Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar, 32Phillips M.A. Rutter W.J. Biochemistry. 1996; 35: 6771-6776Crossref PubMed Scopus (25) Google Scholar) or has only a modest effect (33Gray A.M. Mason A.J. Science. 1990; 247: 1328-1330Crossref PubMed Scopus (221) Google Scholar). The function of mammalian proregions and their ability to function with an extended range of target proteins among protein families have been less well defined. Here we report that the TGD functions in trans to substantially enhance secretion of the truncated protein and of some, but not all, other cytokines. The specific role(s) of proregions as well as their subcellular site of action is diverse. Notably, in another member of the hematopoietic cytokine family, the prolactin proregion is essential for secretion; mutation of the proregion results in accumulation of aggregates in the endoplasmic reticulum (34Haynes R.L. Zheng T. Nicchitta C.V. J. Biol. Chem. 1997; 272: 17126-17133Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar). The proregion of human neutrophil defensin is essential for secretion; the proregion of this protein is anionic (35Valore E.V. Ganz T. Blood. 1992; 79: 1538-1544Crossref PubMed Google Scholar, 59Liu L. Ganz T. Blood. 1995; 85: 1095-1103Crossref PubMed Google Scholar, 60Barker R.L. Gleich G.J. Pease L.R. J. Exp. Med. 1988; 168: 1493-1498Crossref PubMed Scopus (99) Google Scholar), can neutralize the activity of the protein prior to cleavage, and contains a region that is essential for proper subcellular sorting to granule-like vacuoles in cells. Similarly, the proregion of von Willebrand factor (vWF) promotes inter-dimer disulfide bond formation, multimer formation, and targets vWF to storage granules in several in vitro studies (36Journet A.M. Saffaripour S. Wagner D.D. Thromb. Haemostasis. 1993; 70: 1053-1057Crossref PubMed Scopus (39) Google Scholar, 37Journet A.M. Saffaripour S. Cramer E.M. Tenza D. Wagner D.D. Eur. J. Cell. Biol. 1993; 60: 31-41PubMed Google Scholar). These conclusions were confirmed by the finding that a naturally occurring mutation in the proregion of vWF resulted in impaired multimerization and secretion. In the studies reported here, we show that the activity of TGD appears to take place within the endoplasmic reticulum or earlier in the secretory pathway. The cDNA-encoding murine TPO (described previously in Ref. 16Lok S. Kaushansky K. Holly R.D. Kuijper J.L. Lofton-Day C.E. Oort P.J. Grant F.J. Heipel M.D. Burkhead S.K. Kramer J.M. et al.Nature. 1994; 369: 565-568Crossref PubMed Scopus (1042) Google Scholar) was used as a template for PCR-based mutagenesis. To facilitate iodination and purification, PCR-based site-directed mutagenesis was used to add a poly(Tyr) and poly(His) terminus and to mutate an Arg-Arg (Arg-153↓Arg-154) potential cleavage site to Gln-Gln. Oligonucleotides were designed to generate two forms of murine TPO, full-length (TPO 1–335) and a truncated form (TPO 1–152), with a poly(Tyr) and poly(His) carboxyl terminus using a strategy we have described previously (38Matous J.V. Langley K. Kaushansky K. Blood. 1996; 88: 437-444Crossref PubMed Google Scholar). The secretory efficiency (as measured by ELISA) and function (as measured by MTT assay, see below) of TPO 1–335 did not differ from that of wild type mTPO, confirming that mutation of the Arg-Arg site and addition of the poly(Tyr) and poly(His) tail were functionally silent. Site-directed mutagenesis was also used to generate a deletion mutein of TPO in which the entire receptor binding domain (residues 4–174) was removed, termed the TPO glycan domain (TGD); we then modified this construct further by using site-directed mutagenesis to construct TGD variants with a Lys-Asp-Glu-Leu endoplasmic retention signal at the carboxyl terminus (TGD-KDEL) and one lacking the secretory leader (TGD-SL). Fig.1 illustrates the cDNA constructs utilized in this study. Each mTPO construct was cloned into the mammalian expression vector pDX (39Kaushansky K. O'Hara P.J. Hart C.E. Forstrom J.W. Hagen F.S. Biochemistry. 1987; 26: 4861-4867Crossref PubMed Scopus (66) Google Scholar). Cytokine or TGD cDNA expression vectors were co-transfected using LipofectAMINE® (Invitrogen), Lipofectin® (Invitrogen), or calcium phosphate into the rodent fibroblast cell lines BHK with a second plasmid encoding RSV-CAT (chloramphenicol acetyltransferase) to control for transfection efficiency, used at one-tenth the concentration of the cytokine expression vector (40Shoemaker S.G. Hromas R. Kaushansky K. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 9650-9654Crossref PubMed Scopus (76) Google Scholar). Cells were then cultured in Dulbecco's modified Eagle's medium with 2% fetal calf serum and penicillin, streptomycin, and fungizone. In co-transfection experiments with TGD constructs, equal amounts of cytokine cDNA were transfected with TGD cDNA (e.g. 1 μg each for a 2-ml plate). At 24–72 h post-transfection, supernatants were collected and cells counted and lysed. Supernatants were assayed for TPO (see below) and lysates for CAT activity (see below). For all transient transfection experiments, at least two different plasmid preps were utilized. To develop stable transfected cell lines, the TGD encoding cDNA was co-transfected with a plasmid encoding dihydrofolate resistance, at one-tenth the concentration of the cytokine-containing plasmids. Cells were selected and grown in 1 μm methotrexate-containing media, and the supernatant was assayed for TGD secretion and RNA by RT-PCR to confirm the stable expression of TGD. In some experiments cDNA expression vectors were used to produce metabolically labeled proteins. Six hours following transient transfection, cultures were trypsinized and split into two plates to allow metabolic labeling of half the transfected cells, and the remaining cells were left unlabeled in standard culture medium. Eighteen hours following transfection adherent cells were washed, and the supernatant was replaced with Met- and Cys-deficient media (Invitrogen) for 1 h. The media were again replaced with Met- and Cys-deficient media containing 1% dialyzed fetal calf serum, antibiotics, and 50 μCi/ml 35S-labeled Met and Cys (PerkinElmer Life Sciences) and incubated overnight. Supernatants were then collected and clarified, and cells were trypsinized, counted, and lysed for CAT assays. The parallel non-labeled culture supernatant was evaluated for mTPO by ELISA (see below). For experiments using reticulocyte lysates (TNT® SP6 Coupled Reticulocyte Lysate System, Promega), standard protocols supplied by the manufacturer were followed (using RedivueTM35S from Amersham Biosciences for metabolic labeling). The constructs shown in Fig. 1 were cloned into pcDNA3 (Invitrogen) downstream of the Sp6 promoter. Reticulocyte lysate transcripts were electrophoretically size-fractionated on a gradient gel (NOVEXTM, 4–20% acrylamide), which was then dried, exposed to film overnight, and then to a PhosphorImaging screen for 2 h. Two anti-peptide antibodies were generously provided by T. Kato at Kirin Pharmaceuticals; each anti-peptide antibody was directed against a region of the receptor-binding domain of rat TPO. RT1 reacts with a 20-residue region of the putative first helix (9PRLLNKLLRDSYLLHSRLSQ28) and RT2 is directed against a 21-residue segment of the putative AB loop (46FSLGEWKTQTEOSKAQDILGA66). Murine and rat sequences are highly conserved in this region differing at only 2 and 1 residues, respectively (anti-peptide antibodies directed against similar regions of human TPO have been described previously (41Tahara T. Kuwaki T. Matsumoto A. Morita H. Watarai H. Inagaki Y. Ohashi H. Ogami K. Kato T. Stem Cells. 1998; 16: 54-60Crossref PubMed Scopus (14) Google Scholar)). Immunoprecipitation of metabolically labeled tissue culture supernatants was performed using standard protocols. Briefly, 1 ml of metabolically labeled tissue culture supernatant was pre-cleared by incubation with 20 μl of protein A-agarose beads (Santa Cruz Biotechnology) for ½ h; phenylmethylsulfonyl fluoride, leupeptin, and aprotinin were added to diminish proteolysis (as described previously (42Drachman J.G. Griffin J.D. Kaushansky K. J. Biol. Chem. 1995; 270: 4979-4982Abstract Full Text Full Text PDF PubMed Scopus (199) Google Scholar)), and 1% Tween 20 was used to diminish nonspecific binding. Each supernatant was then incubated overnight with 10 μg/ml of the anti-TPO peptide antibody at 4 °C. Protein A-agarose beads were added to precipitate the antibody-TPO complex, and the beads were washed with RIPA buffers as described previously (19Linden H.M. Kaushansky K. Biochemistry. 2000; 39: 3044-3051Crossref PubMed Scopus (35) Google Scholar) and then denatured and eluted from staphylococcus A beads by boiling in a loading buffer with 2% β-mercaptoethanol. The immunoprecipitated TPO forms were size-fractionated by electrophoresis through 10% polyacrylamide gels, soaked in AmplifyTMenhancer (Amersham Biosciences), dried down, and exposed to PhosphorImaging screens overnight. Baf/3 cells transfected with the mouse Mpl receptor were used to assay for TPO activity (16Lok S. Kaushansky K. Holly R.D. Kuijper J.L. Lofton-Day C.E. Oort P.J. Grant F.J. Heipel M.D. Burkhead S.K. Kramer J.M. et al.Nature. 1994; 369: 565-568Crossref PubMed Scopus (1042) Google Scholar). Cells were grown and maintained in IL-3, washed, plated at a concentration of 10,000 cells per well in 100 μl of media in 96-well plates, and incubated for 36 h with serial dilutions of tissue culture supernatants. Wild type mTPO supernatant of known concentration was used to generate a standard curve and determine maximal proliferation of Baf/Mpl cells for each assay. 3[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide, (MTT Sigma) was then added and was followed 5 h later by a lysis buffer. Optical density was measured by using an ELISA plate reader (EL-340 Biotek Instruments) to measure the absorbance at 570–630 nm in order to assess the intracellular conversion of tetrazolium to formazan. Sample activity was determined by comparison to the standard at half-maximal proliferation. The cell line used for these experiments was more sensitive to IL-3 than to TPO; however, as measurement of response to TPO was our desired effect, we have based our results on percentage of mTPO-induced maximal activity. Similarly, MO7e cells, which constitutively express c-Kit and GM-CSF (and a low level of Mpl) receptors and proliferate in response to SCF and GM-CSF, were used to assay for these cytokines. To test SCF activity selectively, Baf/murine Kit receptor cells were also utilized in MTT assays. Baf/EPO receptor cells were used to measure EPO activity, and Baf/3 cells were used to measure IL-3 activity in similar assays. In a similar fashion, megakaryocyte assays were used to quantify SDF-1, as described previously (43Wang Q. Miyakawa Y. Fox N. Kaushansky K. Blood. 2000; 96: 2093-2099Crossref PubMed Google Scholar). The activation of mitogen-activated protein kinase (MAPK) in Baf/Mpl cells was utilized to read out evidence of complete refolding of denatured TPO, in a similar fashion to the assay employed by Baker et al. (44Baker D. Sohl J.L. Agard D.A. Nature. 1992; 356: 263-265Crossref PubMed Scopus (285) Google Scholar) to demonstrate refolding of α-lytic protease. Full-length TPO was generously provided by Dr. Don Foster (Zymogenetics Inc., Seattle, WA), and truncated hormone (containing the amino-terminal 162 residues) was generously provided by Dr. Takashi Kato (Kirin Pharmaceuticals, Takasaki, Japan). Equipotent solutions of truncated and full-length hormone were verified by equal activity in MTT assays using Baf/Mpl cells, as described above. Each protein was then denatured in 6 m guanidine hydrochloride (GdnHCl) at room temperature overnight and heated to 65 °C for 10 min to fully denature the protein (as described previously (45Hamburger J.B. Chen E. Narhi L.O., Wu, G.M. Brems D.N. Proteins. 1998; 32: 495-503Crossref PubMed Scopus (7) Google Scholar)) A roughly 0.8 m solution of the protein was then diluted into cold phosphate-buffered saline with 0.01% bovine serum albumin at room temperature to effectively dilute the protein concentration and GdnHCl 200-fold. The protein was then allowed to refold over a time course of 0–30 min, and aliquots of the refolded protein were then immediately added to starved Baf/Mpl cells, resulting in a final concentration of 0.015 mm GdnHCl. 1 × 106starved Baf/Mpl cells in serum-free media were incubated with diluted protein samples for 2 min in a total volume of 10 ml at 37 °C and then immediately poured into 40 ml of ice-cold phosphate-buffered saline; the cells were pelleted and lysed in freshly prepared lysis buffer (46Miyakawa Y. Oda A. Druker B.J. Kato T. Miyazaki H. Handa M. Ikeda Y. Blood. 1995; 86: 23-27Crossref PubMed Google Scholar) and frozen at 0 °C. Whole cell lysates were size-fractionated by electrophoresis through 8% polyacrylamide gels and transferred to nitrocellulose, blocked with 3% bovine serum albumin, and probed for phosphorylated (activated) MAPK (Cell Signaling antibody number 9101L), and MAPK protein (Cell Signaling antibody number 9109), as we have described previously (47Rojnuckarin P. Miyakawa Y. Fox N.E. Deou J. Daum G. Kaushansky K. J. Biol. Chem. 2001; 276: 41014-41022Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). Repeated optimization assays demonstrated that a minimum of 2 min at 37 °C was necessary for visualization of MAPK phosphorylation and that the addition of trace quantities of GdnHCl to native protein and starved cells did not impair MAPK activation. The proteasome inhibitor ALLN was used to determine whether intracellular protein degradation might account for the reduced secretion of truncated forms of TPO. COS cells were tested for tolerance to ALLN (Calbiochem 108909 Calpain Inhibitor dissolved in Me2SO), and a time course of TPO expression following transient transfection was examined to determine the optimal timing of metabolic labeling and proteasome inhibition. COS cells were transiently transfected as described above, with truncated and full-length TPO in parallel; serum-free, liposome-, and TPO plasmid-containing media were washed off the cells, and they were allowed to recover at 37 °C overnight in media with 2% fetal calf serum. The following day the cells were washed with phosphate-buffered saline and Cys- and Met-deficient media were added with radiolabeled35S as described above. Following a 4-h incubation with labeled culture medium, ALLN or diluent (Me2SO) was then added to the culture to achieve a final concentration of 50 mm, and cells were incubated additionally at 37 °C for 12 h. Supernatants were collected, and cells were harvested and lysed as described above. Similarly, supernatants and cell lysates were analyzed by immunoprecipitation (IP), size fractionation, and PhosphorImaging quantitation as described above. Standard protocols were used to assay CAT activity (40Shoemaker S.G. Hromas R. Kaushansky K. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 9650-9654Crossref PubMed Scopus (76) Google Scholar, 48 ) and thereby correct for transfection efficiency. Cellular lysates were diluted in 0.25 m Tris and incubated with acetyl-CoA and [14C]chloramphenicol. Following ethyl acetate extraction, thin layer chromatography was performed to allow quantitation of chloramphenicol acetylation. If transfection efficiency varied by more than 4-fold between samples, we did not include the results from corresponding supernatants in our analysis. Supernatants were tested in paired samples using a commercial ELISA for mTPO (MTP00, QuantikineTM M from R & D Systems, Minneapolis, MN); in this immunoassay the lower limit is 62.5 pg/ml. The antibodis for this ELISA were also purchased separately and their TPO-binding epitopes were mapped, using purified truncated forms of mTPO, to identify which region(s) of TPO they detect. Both the monoclonal catch antibody and the polyclonal detection antibody detected mTPO by IP and standard Western blot techniques. The monoclonal antibody detected full-length TPO and a truncated TPO form 50 kDa in length but failed to detect shorter forms of TPO, where the polyclonal antibody detected 18-, 20-, 30-, 50-, and 70-kDa forms of TPO, folded and denatured, suggesting that at least part of the epitope of the monoclonal antibody maps to a segment
Referência(s)