Heat Shock Factor 1 Mediates Hemin-induced hsp70 Gene Transcription in K562 Erythroleukemia Cells

Abstract

Transcriptional induction of thehsp70 gene is mediated by heat shock factor 1 (HSF1) rapidly activated upon heat and other stresses. HSF2 has been thought to be responsible for accumulation of HSP70 during hemin-induced differentiation of human K562 erythroleukemia cells because of accompanying acquisition of HSF2 DNA binding activity. However, there has not been any direct evidence for such a functional role of HSF2. The purpose of this study is to clarify the roles of HSF1 and HSF2 in HSP70 induction in hemin-treated K562 cells. We show here that a chimeric polypeptide of HSF2 and GAL4 DNA binding domain (GAL4-BD–HSF2) was unable to induce a GAL4 binding site-containing luciferase reporter gene in response to hemin and that exogenously overproduced HSF2 also failed to increase expression of a heat shock element-containing reporter. On the contrary, expression of a GAL4-BD–HSF1 chimeric protein responded to hemin treatment as well as to heat shock, and transiently overexpressed HSF1 caused hemin-responsive induction of the reporter gene in a dose-dependent manner. These results indicate that HSF1, rather than HSF2, primarily mediates the hemin-induced transcription of the hsp70 gene. Transcriptional induction of thehsp70 gene is mediated by heat shock factor 1 (HSF1) rapidly activated upon heat and other stresses. HSF2 has been thought to be responsible for accumulation of HSP70 during hemin-induced differentiation of human K562 erythroleukemia cells because of accompanying acquisition of HSF2 DNA binding activity. However, there has not been any direct evidence for such a functional role of HSF2. The purpose of this study is to clarify the roles of HSF1 and HSF2 in HSP70 induction in hemin-treated K562 cells. We show here that a chimeric polypeptide of HSF2 and GAL4 DNA binding domain (GAL4-BD–HSF2) was unable to induce a GAL4 binding site-containing luciferase reporter gene in response to hemin and that exogenously overproduced HSF2 also failed to increase expression of a heat shock element-containing reporter. On the contrary, expression of a GAL4-BD–HSF1 chimeric protein responded to hemin treatment as well as to heat shock, and transiently overexpressed HSF1 caused hemin-responsive induction of the reporter gene in a dose-dependent manner. These results indicate that HSF1, rather than HSF2, primarily mediates the hemin-induced transcription of the hsp70 gene. heat shock factor heat shock protein GAL4 DNA binding domain heat shock element phosphate-buffered saline. Cells synthesize increased levels of HSP70 upon various environmental stresses such as elevated temperatures, heavy metals, and amino acid analogs to cope with the deleterious circumstances (1Lindquist S. Craig E.A. Annu. Rev. Genet. 1988; 22: 631-677Crossref PubMed Scopus (4438) Google Scholar). The expression of HSP70 is also increased in mouse embryonal carcinoma cells (2Bensaude O. Morange M. EMBO J. 1983; 2: 173-177Crossref PubMed Scopus (115) Google Scholar), at the two-cell stage of embryogenesis (3Bensaude O. Babinet C. Morange M. Jacob F. Nature. 1983; 305: 331-333Crossref PubMed Scopus (276) Google Scholar), during spermatogenesis (4Zakeri Z.F. Wolgemuth D.J. Mol. Cell. Biol. 1987; 7: 1791-1796Crossref PubMed Scopus (88) Google Scholar), and during S-phase of the cell cycle (5Milarski K.L. Morimoto R.I. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 9517-9521Crossref PubMed Scopus (260) Google Scholar). Heat shock factors (HSFs)1 are responsible for transcription of the hsp70 gene upon these stressful or other physiological conditions, and multiple HSF genes have been isolated in chicken (HSF1, HSF2, and HSF3), mouse (HSF1 and HSF2), and human (HSF1, HSF2, and HSF4) (6Rabindran S.K. Giorgi G. Clos J. Wu C. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 6906-6910Crossref PubMed Scopus (377) Google Scholar, 7Schuetz T.J. Gallo G.J. Sheldon L. Tempst P. Kingston R.E. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 6911-6915Crossref PubMed Scopus (260) Google Scholar, 8Sarge K.D. Zimarino V. Holm K. Wu C. Morimoto R.I. Genes Dev. 1991; 5: 1902-1911Crossref PubMed Scopus (299) Google Scholar, 9Nakai A. Morimoto R.I. Mol. Cell. Biol. 1993; 13: 1983-1997Crossref PubMed Scopus (209) Google Scholar, 10Nakai A. Tanabe M. Kawazoe Y. Inazawa J. Morimoto R.I. Nagata K. Mol. Cell. Biol. 1997; 17: 469-481Crossref PubMed Scopus (288) Google Scholar). Whereas HSF1 and HSF3 are rapidly activated by various stresses and stimulate transcription of the hsp70 gene (11Sarge K.D. Murphy S.P. Morimoto R.I. Mol. Cell. Biol. 1993; 13: 1392-1407Crossref PubMed Scopus (754) Google Scholar, 12Baler R. Dahl G. Voellmy R. Mol. Cell. Biol. 1993; 13: 2486-2496Crossref PubMed Scopus (395) Google Scholar, 13Nakai A. Kawazoe Y. Tanabe M. Nagata K. Morimoto R.I. Mol. Cell. Biol. 1995; 15: 5268-5278Crossref PubMed Scopus (90) Google Scholar, 14Tanabe M. Nakai A. Kawazoe Y. Nagata K. J. Biol. Chem. 1997; 272: 15389-15395Abstract Full Text Full Text PDF PubMed Scopus (71) Google Scholar), it remained to be clarified which of the HSFs is responsible for the hsp70 expression under non-stressed conditions. Human K562 erythroleukemia cells, established from a patient with chronic myelogenous leukemia in blast crisis (15Lozzio C.B. Lozzio B.B. Blood. 1975; 45: 321-334Crossref PubMed Google Scholar), are multipotent stem cells and differentiate to erythroid cells by treatment with hemin or sodium butyrate or hydroxyurea (16Lozzio B.B. Lozzio C.B. Bamberger E.G. Feliu A.S. Proc. Soc. Exp. Biol. Med. 1981; 166: 546-550Crossref PubMed Scopus (220) Google Scholar, 17Andersson L.C. Jokinen M. Gahmberg C.G. Nature. 1979; 278: 364-365Crossref PubMed Scopus (331) Google Scholar, 18Rutherford T. Clegg J.B. Higgs D.R. Jones R.W. Thompson J. Weatherall D.J. Proc. Natl. Acad. Sci. U. S. A. 1981; 78: 348-352Crossref PubMed Scopus (133) Google Scholar). During hemin-mediated erythroid maturation of K562 cells, accumulation of HSP70 (19Singh M.K. Yu J. Nature. 1984; 309: 631-633Crossref PubMed Scopus (51) Google Scholar) and elevated levels of activated HSF2 (20Theodorakis N.G. Zand D.J. Kotzbauer P.T. Williams G.T. Morimoto R.I. Mol. Cell. Biol. 1989; 9: 3166-3173Crossref PubMed Scopus (64) Google Scholar, 21Sistonen L. Sarge K.D. Phillips B. Abravaya K. Morimoto R.I. Mol. Cell. Biol. 1992; 12: 4104-4111Crossref PubMed Scopus (238) Google Scholar) have been observed. Increased expression of HSP70 and constitutive DNA binding of HSF2 have also been reported in mouse spermatogenic cell types (22Sarge K.D. Park-Sarge O.K. Kirby J.D. Mayo K.E. Morimoto R.I. Biol. Reprod. 1994; 50: 1334-1343Crossref PubMed Scopus (146) Google Scholar), embryonal carcinoma cells (23Murphy S.P. Gorzowski J.J. Sarge K.D. Phillips B. Mol. Cell. Biol. 1994; 14: 5309-5317Crossref PubMed Scopus (70) Google Scholar), and postimplantation embryos from E8.5 to E15.5 (24Rallu M. Loones M. Lallemand Y. Morimoto R. Morange M. Mezger V. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 2392-2397Crossref PubMed Scopus (109) Google Scholar). However, whether HSF2 mediates the transcriptional activation of the hsp70 gene under these non-stress conditions has been controversial for lack of direct evidence. We recently analyzed the functional domains of human HSF2 in K562 cells using chimeric constructs fused to the GAL4 DNA binding domain (GAL4-BD) and identified the transactivation domains and negative regulatory motifs (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar). Although chimeric constructs of HSF1 fused to GAL4-BD had been shown to respond to heat shock by activating transcription of the heat shock element (HSE)-containing reporter genes (26Zuo J. Rungger D. Voellmy R. Mol. Cell. Biol. 1995; 15: 4319-4330Crossref PubMed Scopus (207) Google Scholar, 27Shi Y. Kroeger P.E. Morimoto R.I. Mol. Cell. Biol. 1995; 15: 4309-4318Crossref PubMed Scopus (102) Google Scholar, 28Newton E.M. Knauf U. Green M. Kingston R.E. Mol. Cell. Biol. 1996; 16: 839-846Crossref PubMed Scopus (116) Google Scholar), all of the GAL4-BD–HSF2 chimeras failed to respond to hemin treatment. Such an unexpected behavior of HSF2 prompted us to investigate which HSF mediates the increased expression of HSP70 in hemin-treated K562 cells. Here, we report that, in K562 cells (ATCC CCL243), HSF1 but not HSF2 responds to hemin to activate transcription of the hsp70gene. GAL4-BD–HSF1 chimeric protein stimulated expression of a luciferase reporter gene containing GAL4 binding sites upon not only heat shock but also hemin treatment. Transiently overexpressed HSF1 increased expression of an HSE-containing reporter in response to hemin in a dose-dependent manner, but transient expression of HSF2 did not show any significant effects on the hemin response. Moreover, HSF2-overproducing stable transformants lacked hemin-responsive expression of HSP70 despite constitutive HSF2 DNA binding. These results suggest that HSF1 primarily mediates transcription of the hsp70 gene during hemin-induced differentiation of K562 cells. The fusion gene of HSF2N96 or HSF1N101 with GAL4-BD was inserted into a mammalian expression vector pCAGGS (29Niwa H. Yamamura K. Miyazaki J. Gene (Amst.). 1991; 108: 193-199Crossref PubMed Scopus (4617) Google Scholar) as described previously (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar). To construct HSF1+2N101, residues 215 to 378 of HSF1 were replaced with residues 204 to 354 of HSF2 by polymerase chain reaction. In HSF2+1N96, residues 210 to 321 of HSF1 were substituted for residues 199 to 327 of HSF2. HSF2 and HSF1 expression plasmids, pCAHSF2 and pCAHSF1, were constructed by ligation into pCAGGS (29Niwa H. Yamamura K. Miyazaki J. Gene (Amst.). 1991; 108: 193-199Crossref PubMed Scopus (4617) Google Scholar) of the DNA fragments encoding human HSF1 and HSF2 prepared from phHSF2–1 (provided by R. E. Kingston) and phHSF1 (30Yoshima T. Yura T. Yanagi H. Biochem. Biophys. Res. Commun. 1997; 240: 228-233Crossref PubMed Scopus (12) Google Scholar), respectively. pCXHSF2 was constructed by insertion of the HSF2 fragment into pCXN2 (29Niwa H. Yamamura K. Miyazaki J. Gene (Amst.). 1991; 108: 193-199Crossref PubMed Scopus (4617) Google Scholar) to generate HSF2-overproducing stable cell lines. The firefly luciferase reporter plasmid pGLHSE3 was constructed by insertion into pGL3-promoter vector (Promega) of three tandem repeats of an HSE-containing synthetic DNA, whose sequence was identical to that of the probe used in gel shift analysis (see below). GAL4 binding site-containing luciferase reporter plasmid pGLG4E5 was described previously (30Yoshima T. Yura T. Yanagi H. Biochem. Biophys. Res. Commun. 1997; 240: 228-233Crossref PubMed Scopus (12) Google Scholar). Renilla luciferase expression plasmid pRLSV40 was purchased from Promega. K562 cells (ATCC CCL243) were cultured in RPMI 1640 medium containing 10% fetal bovine serum in 5% CO2 at 37 °C. Transient transfection of K562 cells was done by electroporation. 5 × 106cells were washed twice with phosphate-buffered saline (PBS) and suspended in 0.4 ml of PBS containing 30 μg of plasmids. After incubating for 10 min at room temperature, the cells were exposed to an electric pulse (300 V/cm, 500 microfarads) by using Gene Pulser (Bio-Rad). The electroporated cells were left at room temperature for 10 min and then incubated in RPMI 1640 medium containing 10% fetal bovine serum in 5% CO2 at 37 °C for 48 h. Cells were harvested, washed twice with PBS, divided into two aliquots for luciferase assay and for preparation of cell extracts, and stored at –80 °C. Luciferase assays were performed by using Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instruction. K562 cells were transfected with 20 μg of linearized pCXHSF2 DNA as described above. Cells were cultured for 2 days at 37 °C, and neomycin-resistant cells were selected in medium containing 300 μg/ml of G418 for 4 days. Thereafter, the culture was diluted, and the selection was continued further for 2 weeks to obtain single cell-derived clones. Cell pellets were frozen at –80 °C and suspended in 20 mm HEPES containing 1.5 mm MgCl2, 0.2 mm EDTA, 0.5 mm dithiothreitol, 25% glycerol, 1 mmphenylmethylsulfonyl fluoride, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, and 10 μg/ml aprotinin. After addition of NaCl to a final concentration of 0.42 m, the lysates were set on ice for 5 min and centrifuged for 15 min at 100,000 × g at 4 °C. The supernatants were used as whole cell extracts. Protein concentrations were determined by a protein assay kit (Bio-Rad) using bovine serum albumin as a standard. Total RNA from K562 cells was prepared by using Isogen (Nippon Gene) and separated on a 1% agarose gel containing formaldehyde. The RNA was transferred onto a nylon membrane (DuPont NEN), hybridized with 32P-labeled humanhsp70 DNA DNA to prepared from DNA by polymerase chain or washed in 0.2 × at °C, by and with a cell μg of were with 5 of containing of the human hsp70 to Williams G.T. Morimoto R.I. J. Kingston R.E. R. Mol. Cell. Biol. 1987; 7: PubMed Scopus Google Scholar) in of mm mm 1 mm EDTA, 0.5 5% containing 1 μg of The was incubated for 20 min at room temperature, and 15 was onto a gel in × After at for 2 at 4 °C, the gel was on and exposed to shift assay using a probe containing a single GAL4 binding was described previously (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar). chicken HSF2 and or chicken HSF1 by A. or human HSF1 were cell μg of were to gel and were onto The were with 5% in PBS containing 20 for 1 at room temperature and incubated with in containing for 1 at room After with the were incubated with for 1 at room with were using and with of binding of HSF2 and transcription induction of the hsp70 gene were in hemin-treated K562 cells by and gel shift analysis levels of hsp70 increased and by and hemin treatment for and by heat shock °C for 1 1 A, K562 cells were with accumulation of hsp70 was at and the was at and The of increased and with hemin at h. The of binding was in the cells with hemin for 1 4 and and the induction of the binding was well with the increase of hsp70 These results were with the reported previously (20Theodorakis N.G. Zand D.J. Kotzbauer P.T. Williams G.T. Morimoto R.I. Mol. Cell. Biol. 1989; 9: 3166-3173Crossref PubMed Scopus (64) Google Scholar, 21Sistonen L. Sarge K.D. Phillips B. Abravaya K. Morimoto R.I. Mol. Cell. Biol. 1992; 12: 4104-4111Crossref PubMed Scopus (238) Google Scholar). has been reported that the of HSF2 is during hemin treatment L. Sarge K.D. Morimoto R.I. Mol. Cell. Biol. 1994; 14: PubMed Scopus Google Scholar, S. L. H. Sarge K.D. L. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar). we also increase of the HSF2 in K562 cells upon treatment with 30 and hemin for 1 Moreover, increase of of HSF1 was in the hemin-treated the of to on the hemin concentration 1 as has been with HSF1 upon heat shock 5 and K.D. Murphy S.P. Morimoto R.I. Mol. Cell. Biol. 1993; 13: 1392-1407Crossref PubMed Scopus (754) Google Scholar). The was and of hemin treatment has been shown to be to HSF1 J.J. M. Morimoto R.I. J. Biol. Chem. 1996; Full Text Full Text PDF PubMed Scopus Google Scholar, Voellmy R. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar), these results the that hemin HSF1 to hsp70 To which HSF is in the hemin-induced binding we cell with HSF1 or HSF2 and performed gel shift shown in only a of the binding by 30 or hemin were by HSF2 and under the conditions that caused for overexpressed but not for heat On the other all in the hemin-treated cell were by HSF1 and as well as in the cell the to be as with that in the cell and These results that the hemin-induced binding of HSF1 and HSF2, the We then which HSF is in the transcriptional activation in hemin-treated K562 cells by luciferase reporter HSF2N96 and HSF1N101 have of the of HSF2 and HSF1, for DNA binding domains that were replaced by GAL4-BD (30Yoshima T. Yura T. Yanagi H. Biochem. Biophys. Res. Commun. 1997; 240: 228-233Crossref PubMed Scopus (12) Google Scholar). to HSF1N101 had been shown to respond to heat shock and activate transcription of reporter genes (26Zuo J. Rungger D. Voellmy R. Mol. Cell. Biol. 1995; 15: 4319-4330Crossref PubMed Scopus (207) Google Scholar, 27Shi Y. Kroeger P.E. Morimoto R.I. Mol. Cell. Biol. 1995; 15: 4309-4318Crossref PubMed Scopus (102) Google Scholar, M. T.J. Kingston R.E. Mol. Cell. Biol. 1995; 15: PubMed Scopus Google Scholar). The heat shock regulatory domain of HSF1 was substituted for the of HSF2 and to chimeric constructs DNA binding of these constructs in K562 cells was by gel shift assay using a single GAL4 binding as a probe The transactivation of construct HSF2N96 was under conditions and was not by to hemin or heat shock A, The chimeric construct also response to hemin or heat shock is that containing the heat shock regulatory domain of HSF1, which had been shown to be to heat to a transactivation domain E.M. Knauf U. Green M. Kingston R.E. Mol. Cell. Biol. 1996; 16: 839-846Crossref PubMed Scopus (116) Google Scholar), failed to activate transcription upon heat constructs the negative regulatory acid that the transactivation domain of HSF2 (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar), these results suggest that the negative of HSF2 is functional in or heat K562 cells of the regulatory domain of On the other construct HSF1N101 elevated the reporter expression and in response to hemin and heat shock, construct the heat shock regulatory domain was and in constitutive activation of the reporter HSF2 did not show any response to hemin and heat shock, HSF1 responded well to at under the conditions that GAL4-BD was used as a DNA binding To further the of HSF1 in hemin-induced transcription of thehsp70 HSF1 or HSF2 was transiently in K562 cells, and expression of the HSE-containing luciferase gene was shown in 4 A, the reporter response of the cells was to that of the cells, which the response mediated by In transiently HSF1 the reporter response not only to heat shock but to hemin treatment and as whole cell prepared from the transfected cells were by gel shift the binding of transiently HSF1 was increased by treatment with hemin or heat shock 4 4 and such increase in binding was with the cells 4 5 and The transient expression of HSF2 or HSF1 was by increased of protein in the transfected cells 4 These results also suggest that HSF1, rather than HSF2, is activated to stimulate transcription of thehsp70 gene in hemin-treated K562 cells. we failed to hemin of the K562 cells that had been transiently transfected with expression for fusion protein or HSF2 and stable HSF2 cell were established to the that the expression levels were in transient expression The HSF2 expression levels in the stable used were and than that of the K562 cells 5 The overexpressed HSF2 in constitutive binding 5 However, the of hsp70 and were not elevated in these cell 5 A, Moreover, the of HSF DNA binding increased further by hemin treatment 5 the of of transcription was not than with the cells 5 A, On the contrary, of HSF2 to be rather to hemin-induced activation of not only did the of HSF2 to to the hemin-induced activation of hsp70 but it was that it transcription by with HSF1 in binding to on the hsp70 We have shown that HSF1 is primarily in transcriptional activation of the hsp70 gene in hemin-treated K562 cells. has been thought that HSF2 is the transcription factor responsible for hsp70 expression during hemin-induced differentiation because DNA binding of HSF2 upon hemin treatment. However, transiently overexpressed GAL4-BD–HSF2 chimeric protein or HSF2 failed to respond to hemin and activate transcription of the reporter gene containing a GAL4 binding or 4 in cells, the or hemin-induced levels of hsp70 transcription was not than in the cells despite the constitutive HSF2 binding to 2 and On the other GAL4-BD–HSF1 chimeric protein responded to heat shock and hemin treatment and stimulated expression of the reporter gene Transiently overexpressed HSF1 also the hemin-induced expression of the HSE-containing luciferase reporter in a dose-dependent 4 In the of hemin-induced binding was shown to be HSF1 which was as in cells 1 These results suggest that HSF1 but not HSF2 primarily mediates the hemin-induced transcriptional activation of thehsp70 gene in K562 cells. The results here are with the L. Sarge K.D. Phillips B. Abravaya K. Morimoto R.I. Mol. Cell. Biol. 1992; 12: 4104-4111Crossref PubMed Scopus (238) Google Scholar) reported that the hemin-induced binding in gel shift analysis with HSF2 but are only with HSF1 also that of which had been shown to as a transcription Park-Sarge Sarge K.D. Mol. Cell. Biol. 1995; 15: PubMed Scopus Google Scholar), hemin-induced heat shock gene expression S. L. H. Sarge K.D. L. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar). be that HSF2 we used here was but not which had been reported to hemin-induced expression of HSP70 S. L. H. Sarge K.D. L. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar). These be on the of of K562 cells. the three of K562 cells S.K. A. Exp. 1983; Google Scholar), we used which has a to hemin to The HSF to hemin K562 cells in L. Sarge K.D. Morimoto R.I. Mol. Cell. Biol. 1994; 14: PubMed Scopus Google Scholar) also but significant and of HSF1 upon hemin it is that cell was to activate HSF1 in response to The activated HSF1 be to stimulate hsp70 In any the that hsp70 transcription not increase in K562 cells despite constitutive DNA binding 5 and S. L. H. Sarge K.D. L. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar) suggest that the acquisition of HSF2 DNA binding is and that an regulatory is to activate transcription of heat shock genes as in the of HSF1 S.K. Kingston R.E. Mol. Cell. Biol. PubMed Scopus Google Scholar, S.K. Mol. Cell. Biol. 1991; PubMed Scopus Google Scholar, L. Morimoto R.I. 1992; PubMed Scopus Google Scholar, S.K. Res. 1993; Google Scholar). is that the cell we used the to the DNA binding of HSF2 to a HSF2 not to of thehsp70 in K562 cell as in cells (23Murphy S.P. Gorzowski J.J. Sarge K.D. Phillips B. Mol. Cell. Biol. 1994; 14: 5309-5317Crossref PubMed Scopus (70) Google Scholar). In any HSF2 activation not be during differentiation of K562 cells because was in conditions not and induce HSF1 and hsp70 transcription M. Wu J. Biol. Chem. 1996; Full Text Full Text PDF PubMed Scopus Google Scholar), that accumulation of that are by hsp70 transcription activation of hemin also at J.D. J. Biol. Chem. Full Text PDF PubMed Google Proc. Natl. Acad. Sci. U. S. A. 1981; 78: PubMed Scopus Google Scholar), which is with we to induce differentiation of K562 cells. that hemin activate HSF1 the that was by hemin with a to that of HSP70 induction be by the that as well as are by various S.K. J. Biol. Chem. 1997; 272: Full Text Full Text PDF PubMed Scopus Google Scholar). We that the HSP70 induction in hemin-treated K562 cells is primarily mediated by HSF1, which is activated upon accumulation of from of activity. Although we failed to any of HSF2 in hemin-induced hsp70 transcription in K562 cells, it that HSF2 roles during and the of HSF2 to activate transcription from has also been reported with other embryonal carcinoma cells (23Murphy S.P. Gorzowski J.J. Sarge K.D. Phillips B. Mol. Cell. Biol. 1994; 14: 5309-5317Crossref PubMed Scopus (70) Google Scholar) and embryos (24Rallu M. Loones M. Lallemand Y. Morimoto R. Morange M. Mezger V. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 2392-2397Crossref PubMed Scopus (109) Google Scholar), or conditions be for HSF2 transactivation We recently that HSF2 is in K562 cells by with the negative regulatory to the transactivation domain of HSF2 (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar). The gel shift that the are to with a DNA binding of HSF2 (25Yoshima T. Yura T. Yanagi H. Nucleic Acids Res. 1998; 26: 2580-2585Crossref PubMed Scopus (14) Google Scholar). of this and other a in the of HSF2 We are to Nakai for and of and E. Kingston and Miyazaki for plasmids. We also and for and and for