Hsp90 (cid:1) Recruited by Sp1 Is Important for Transcription of 12( S )-Lipoxygenase in A431 Cells *

Sp1 is a basic transcriptional factor that binds to the GC-rich region in the promoter of the target gene. It is involved in transcription of numerous genes by recruiting transcriptional factors to the promotersoftargetgenes.Inthisstudy,wefound invivo and invitro that Hsp90 (cid:1) was recruited to the GC-rich region of the 12( S )-li-poxygenase promoter through interaction with Sp1 in A431 cells by employing DNA affinity immunoprecipitation assay and chromatin immunoprecipitation assay. When Hsp90 (cid:1) was inhibited by geldanamycin (GA, a specific inhibitor of the Hsp90 family) or by siRNA of Hsp90 (cid:1) (to block its activity or to knockdown protein levels), respectively, luciferase activity (driven by the 12( S )-lipoxy-genase promoter) and both mRNA and protein levels of 12( S )-li-poxygenase were reduced significantly in cells. In addition, the effect of GA was abolished when the Sp1 binding sites of 12( S )-lipoxygenase were mutated in A431 cells. Interestingly, binding of Sp1 to the 12( S )-lipoxygenase promoter was also decreased upon GA treatment in cells. In conclusion, our results indicate that Sp1 interacts with Hsp90 (cid:1) to recruit it to the promoter of 12( S )-lipoxy-genase and then to regulate gene transcription by modulating the

Sp1 is one of the first transcription factors purified and cloned from mammalian cells (1,2). It can recognize and specifically bind to GC-rich sites within the simian virus 40 promoter via three Cys 2 His 2 zinc finger motifs localized at its C-terminal region to regulate the transcription of the target genes (3,4). In addition to the zinc finger domain of the C-terminal region, the N-terminal regions of the Sp1 are much more variable and contain transcriptional activation or repression domains (5,6). Sp1 is generally considered as a factor that primarily determines the core activity of the promoter by direct interaction with other factors of the basal transcriptional machinery and by cooperation with several transcriptional activators such as CRSP, p300/CBP, steroidogenic factor-1 (SF-1), vitamin D3 receptor, and TAFII130 (7)(8)(9)(10)(11)(12). Recent studies reveal that both DNA binding ability and transactivational activity of Sp1 may be influenced by the post-translational modification of Sp1 such as phosphorylation, glycosylation, and acetylation (13,14). Previous studies indicate that Sp1 is phosphorylated by casein kinase II, which has been reported to repress Sp1 activity, and by DNA-dependent protein kinase and CDKII, to positively regulate the transactivity and DNA binding affinity of Sp1 (15)(16)(17)(18). In addition, glycosylation of Sp1 was found to regulate the proteasome-dependent degradation, and it is acetylated to regulate the DNA binding affinity or transactivation (19 -24). Therefore, post-translational modification on Sp1, because of interaction with other factors, may play an important role in the regulation of Sp1 activity.
Hsp90, a constituent molecular chaperone, is an abundant protein, comprising 2% of total cellular proteins under non-stress conditions. It is essential for numerous cellular proteins that regulate signal transduction such as transcription factors, protein kinases, and nitric-oxide synthase, and involved in various cellular processes, such as cell proliferation, differentiation, and apoptosis (25)(26)(27)(28)(29). Unlike Hsp70, Hsp90 does not act generally in nascent protein folding, and instead, binds to the client proteins to stabilize the folding of proteins (30, 31). The crystal structure of Hsp90 reveals that the N-terminal domain of Hsp90 binds ATP, which is consistent with the observation that ATP hydrolysis is required for conformational changes involved in refolding protein substrates or client proteins of Hsp90. Geldanamycin (GA), 2 a benzoquinone ansamycin, and radicicol, a macrocyclic anti-fungal antibiotic, can compete the ATP binding site on Hsp90 to inhibit its activity (32,33). Hsp90 resides mostly in the cytoplasm to form the main functional component of an important cytoplasmic chaperone complex (34). However, it has been also found inside the nucleus and outside the cells in unstressed or stressed cells and cancer cells (35)(36)(37)(38). Most studies about nuclear Hsp90 focus on how the Hsp90 shifts glucocorticoid receptor (GR) into the nucleus and regulates nuclear retention (38,39). However, whether the Hsp90 has additional functions within the nucleus is not very clear.
Arachidonate 12(S)-lipoxygenase (arachidonate: oxygen 12-oxidoreductase; EC 1.13.11.31) in the platelet was the first mammalian lipoxygenase discovered (40). It catalyzes the transformation of arachidonic acid into 12(S)-hydroperoxyeicosatetraenoic acid, followed by conversion to 12(S)-HETE. 12(S)-HETE plays a significant role in the pathogenesis of some epidermal and epithelial inflammation. A markedly elevated 12(S)-HETE level was found in psoriatic plaque, whereas the level of prostaglandins E 2 and F 2␣ was only minimally elevated (41,42). Therefore, high levels of 12(S)-HETE may contribute to the inflammatory changes and the abnormal epidermal hyperproliferation in the development of a psoriatic plaque. In terms of promoter activity, the two simian virus 40 promoter factor 1 (Sp1) binding sites residing at Ϫ158 to Ϫ150 bp and Ϫ123 to Ϫ114 bp were found to be required for the transcription of the 12(S)-lipoxygenase (43). Under normal conditions, Sp1 recruited to the promoter of 12(S)-lipoxygenase is responsible for its basal transcription. Upon induction with phorbol 12-myristate 13-acetate (PMA) or epidermal growth factor (EGF), c-Jun can be recruited to the promoter through Sp1 to regulate transcription of 12(S)-lipoxygenase (43,44).
In this study, we first found that Sp1 could recruit Hsp90␣ to the GC-rich region of 12(S)-lipoxygenase promoter in A431 cells. Interestingly, a specific inhibition of Hsp90 by GA blocked Sp1 activity and siRNA of Hsp90␣ resulted in knockdown of the Hsp90 level, inhibiting the transcriptional activity of 12(S)-lipoxygenase in A431 cells. Furthermore, the DNA binding affinity of Sp1 was inhibited in the presence of GA.
Preparation of Nuclear Extracts-Cells were washed twice with PBS and scraped in 10 ml of PBS containing 0.5 mM phenylmethylsulfonyl fluoride. Cells were collected by centrifuging at 1200 ϫ g for 5 min and then resuspended in 400 l of buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 1.5 mM MgCl 2 , 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM DTT, 2 g/ml leupeptin, and 2 g/ml pepstatin). After staying on ice for 10 min, samples were vortexed for 10 s and sucked back and forth six times using 25-gauge needles. Nuclei were pelleted by centrifugation at 12,000 ϫ g for 30 s. The pellets were resuspended in 100 l of buffer C (20 mM HEPES, pH 7.9, 420 mM NaCl, 0.2 mM EDTA, 25% glycerol, 1.5 mM MgCl 2 , 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM DTT, 2 g/ml leupeptin, and 2 g/ml pepstatin) and put on ice for 20 min. Finally, nuclear extracts were prepared by centrifugation at 12,000 ϫ g for 3 min to remove the pellet and stored at Ϫ80°C until use.
Immunoprecipitation-Cell nuclear extracts were prepared, and the protein concentration was determined using a bicinchoninic acid protein assay kit (Pierce), and an equal amount of protein was used in each experiment. Nuclear extracts were preincubated with protein A/G-Sepharose for 1 h at 4°C and centrifuged to remove the pellets. The supernatants were transferred to new tubes and incubated with either anti-Hsp90 or anti-Sp1 antibodies at a dilution of 1:200 for 12 h at 4°C. The immunoprecipitated pellets were subsequently incubated with protein A/G-Sepharose, washed five times with lysis buffer, and separated on an SDS-7% polyacrylamide gel. After electrophoresis, the gels were processed for immunoblot analysis.
Transfection and Reporter Gene Assay-A431 cells (2.5 ϫ 10 5 ) were seeded on 3.5-cm dishes and reached 40 -50% confluent on the day of transfection. Cells were transfected with plasmids by lipofection using Lipofectamine 2000 according to the manufacturer's instruction with a slight modification. For use in transfection, 1 g of pXP-1 or pXLO7-1 luciferase plasmids was combined with 1 l of Lipofectamine 2000 in 200 l of Opti-MEM medium without serum, and incubated at room temperature for 30 min. Cells were transfected by changing the medium with 2 ml of Opti-MEM medium containing the plasmids and Lipofectamine 2000, and then incubating at 37°C in 5% CO 2 for 6 h. After change of Opti-MEM medium to 2 ml of fresh medium containing 10% fetal bovine serum, cells were incubated for an additional 18 h. The luciferase activity in the cell lysate was determined as described previously (44).
DNA Affinity Precipitation Assay (DAPA)-The oligonucleotide, 5Ј-TTTGGGCTAGTCTGGGGCGGGG-3Ј, localized Ϫ171 to Ϫ150 bp within the promoter of 12(S)-lipoxygenase and its mutant, 5Ј-TTT-GGGCTAGTCTAAAACAAAA-3Ј, were biotinylated at the 5Ј-termini, and then annealed with their complementary strands, respectively. The DNA affinity precipitation assay was performed by incubating 2 g of biotinylated DNA probe with 300 g of nuclear extract that was precleaned with 10 l of streptavidin-agarose beads and 1 g of poly(dI-dC) for 1 h, and then incubated with 20 l of streptavidin-agarose in binding buffer (1 g of poly(dI-dC), 20 mM HEPES, pH 7.9, 0.1 mM KCl, 2 mM MgCl 2 , 15 mM NaCl 0.2 mM EDTA, 1 mM DTT, and 10% (v/v) glycerol) for 1 h. Beads were collected and washed with the binding buffer containing 0.5% Nonidet P-40 three times. Proteins bound to the beads were eluted with 2ϫ sample buffer and separated by 7% SDS-PAGE for immunoblot analysis, and by 8 -20% gradient SDS-PAGE for silver staining.
Chromatin Immunoprecipitation (ChIP)-Assays were performed as described previously (45). Cells (1 ϫ 10 8 ) were cross-linked using 0.5% formaldehyde in PBS for 15 min at room temperature. The cells were washed three times with PBS, and the cell lysate was then collected with lysis buffer. The chromatin was fragmented by sonication to an average size of 500 bp. The samples were precleaned with 1 g of poly(dI-dC) and 10 l of protein A/G-agarose for 1 h, and then immunoprecipitated with 5 g of IgG, anti-Sp1, anti-Hsp90, anti-TFIID (TBP), anti-acetyl H3, or anti-actin antibodies. After 12 h of incubation, the samples were incubated with 20 l of protein A/G-agarose for 1 h, respectively. After six washes, bound proteins were eluted with TE buffer containing 1% SDS. Cross-links were reversed at 65°C for 12 h, and proteins were digested with proteinase K (0.5 mg/ml) for 2 h at 50°C. DNA was purified by phenol-chloroform extraction and ethanol precipitation. Immunoprecipitated DNA was analyzed by PCR and Southern blotting. The primer sequences for PCR were as follows: 5Ј-TTTGGGCTAGTCT-GGGGCGGGG-3Ј and 5Ј-GGCGCCCCCCAGCAGCTTAGGC-3Ј and those for Southern blotting were as follows: 5Ј-AGACGGTCCTT-TAAAGGTTGGAAGCC-3Ј and 5Ј-GGCTTCCAACCTTTAAAGG-ACCGTCT-3Ј.
Recombinant Retrovirus-AmphoPack-293 cells (2.5 ϫ 10 6 ) were seeded and then cultured overnight for retrovirus packaging. pSM2 or pSM2-shHSP90 at a dose of 1 g was transfected into the cells with Lipofectamine 2000, and then incubated with fresh medium for 3 days (46). The supernatant was collected and concentrated. The titer was 2 ϫ 10 6 per ml, and the viral particles were stocked at Ϫ80°C.
In Vitro Transcription Assay-The in vitro transcription assay was carried out at 30°C for 1 h in 40 l of reaction mixture containing 12 mM HEPES (pH 7.9), 12% glycerol, 80 mM KCl, 4 mM MgCl 2 , 0.6 mM dithiothreitol, and 0.3 mM EDTA. Nucleotide concentrations were 600 M for ATP, GTP, and CTP and 30 M for UTP containing 5 Ci of ␣-[ 32 P]UTP. The reaction mixture included 10 g of nuclear extract and 1 g of template DNA with or without RNase i to inhibit the RNase activity. The transcripts were extracted once with phenol and once with phenol-chloroform, precipitated with ethanol, and analyzed by electrophoresis on a 5% polyacrylamide-7 M urea gel. Finally, the gel was dried, and autoradiography was then performed.
Two-dimensional Electrophoresis-Two-dimensional electrophoresis was performed using isoelectric focusing (IEF) as the first dimension (IPGphor, Amersham Biosciences), using an immobilized pH gradient strip (IPG strip, 18-cm long, linear pH 4 -7), followed by second-dimensional SDS-PAGE (Bio-Rad Protean II xi cell). The quantity of protein loaded onto the two-dimensional gel was 50 g. IEF was run following a stepwise incremental voltage program: 30 V for 16 h, 500 V for 1 h, 1000 V for 1 h, and 8000 V for 4 h, with a total power of 34 kV-h. After IEF, the strips were subjected to two-step equilibration in equilibration buffers containing 6 M urea, 30% glycerol, 2% SDS, and 50 mM Tris-HCl, pH 8.8 with 1% DTT w/v for the first step, and 2.5% (w/v) iodoacetamide for the second step. The strips were then transferred onto the second-dimensional SDS-PAGE and run on 1.0-mm thick 8 -16% gradient polyacrylamide gels at 10°C. The gels were fixed for silver staining.
Protein Identification by Mass Spectrometry Analysis-The protein spot of interest on the two-dimensional electrophoresis was excised, ground manually, and transferred into a siliconized 0.5-ml microcentrifuge tube for further identification by mass spectrometry analysis. The in-gel digestion, mass spectrometry analysis, and data base search were performed as described previously (47).
Immunofluorescence Microscopic Analysis-A431 cells were seeded on round glass slides in 24-well plates overnight. The cells were fixed in 4% paraformaldehyde at 4°C for 20 min. Cells were rinsed in PBS three times and permeabilized in 1% Triton X-100 for 5 min and in 0.5% Tween 20 for 15 min. These cells were blocked with 1% bovine serum albumin-PBS for 60 min at 25°C and incubated with antibody against Hsp90 at a dilution of 1:500 for 1 h, with FITC-conjugated donkey anti-mouse immunoglobulin G and antibodies against Sp1 at a dilution of 1:250 for 1 h, and with Cy5-conjugated donkey anti-rabbit immunoglobulin G for 1 h. These samples were washed with PBS, mounted in 50% glycerol, and analyzed using a Leica TCS SP2 confocal microscope.

RESULTS
Hsp90␣ Was Recruited by Sp1 to the Promoter of 12(S)-Lipoxygenase in A431 Cells-Sp1 acts as an anchor protein to recruit other transcriptional factors to carry out the transcriptional mechanism (43). Therefore, a biotinylated DNA fragment localized at the Ϫ171 to Ϫ150 bp region of the 12(S)-lipoxygenase promoter containing Sp1 binding sites was used as a probe to study the factors that could be recruited to the promoter by Sp1. The samples were analyzed by immunoblotting with anti-Sp1 antibodies and shown in Fig. 1A. This indicated that Sp1 could bind to the GC-rich region of 12(S)-lipoxygenase, but not to the mGCrich region in which the sequence was modified from 5Ј-TTTGGGCT-AGTCTGGGGCGGGG-3Ј to 5Ј-TTTGGGCTAGTCTAAAACAAA-A-3Ј (Fig. 1A). These samples were subsequently analyzed by twodimensional SDS-PAGE, revealing that more proteins were recruited to the GC-rich probe, but not to the mGC-rich probe (Fig. 1B). The proteins on the gel were collected and analyzed employing MALDI-TOF MS to verify their identities. Among those, Hsp90␣ was clearly identified (Fig. 1C). ChIP assay was carried out to mimic the in vivo recruitment of Hsp90␣ to the promoter of 12(S)-lipoxygenase. The results revealed that GC-rich region of 12(S)-lipoxygenase promoter could be pulled down by anti-Sp1 and anti-Hsp90␣ antibodies (Fig. 1D). These results indicated that the Hsp90␣ could be recruited to the promoter of 12(S)-lipoxygenase by Sp1 in A431 cells. Further, immunoprecipitation with anti-Sp1 antibodies followed by immunoblotting with anti-Hsp90␣ and anti-Sp1 antibodies was performed. As shown in Fig. 2A, Hsp90␣ could be immunoprecipitated by anti-Sp1 antibodies. In addition, we also expressed and purified the GST-Sp1 protein from Escherichia coli and used it to pull-down interacting proteins residing in the nuclear extract. Immunoblots with anti-Hsp90␣ and anti-GST antibodies revealed that GST-Sp1 could pull-down the Hsp90␣, but the GST protein alone could not (Fig. 2B). To investigate whether Sp1 and Hsp90␣ interacted directly, both Hsp90␣ and GST-Sp1 were expressed and purified for in vitro interaction assay. These results revealed that no direct interaction between Sp1 and Hsp90␣ occurred (Fig. 2C). In addition, cells were also observed with confocal microscopy to study the co-localization between the Sp1 and Hsp90␣ (Fig. 2D). The result revealed that a little Hsp90␣ was inside the nucleus, and it was colocalized with the Sp1. In contrast, tubulin ␣, which has been reported to be absent in the nucleus, was used as a negative control. Taken together, these results indicated that Hsp90␣ interacted indirectly with Sp1, and these two proteins were recruited together to the nucleus to bind to the promoter of 12(S)-lipoxygenase in A431 cells.
GA Inhibited the Transcriptional Activity of 12(S)-Lipoxygenase-To investigate the effect of the Hsp90␣ on the transcription of the target gene, A431 cells were treated with GA to inhibit the function of Hsp90␣ by competing with ATP binding at the ATP binding site. The protein synthesis rate, mRNA level, and luciferase activity driven by 12(S)-lipoxygenase promoter (Fig. 3C) in cells were then analyzed. A431 cells were treated with various concentration of GA for 18 h and then labeled with [ 35 S]methionine for 1 h. IKK␣, which can be inhibited by GA treat-ment was used as a positive control, whereas Sp1 was used as a negative control. The 12(S)-lipoxygenase, IKK␣, and Sp1 levels were then analyzed with immunoprecipitation by anti-12(S)-lipoxygenase, anti-IKK␣, and anti-Sp1 antibodies, respectively (48,49). After autoradiography, the results revealed that synthesis rate of 12(S)-lipoxygenase and IKK␣ were reduced in the presence of GA at a concentration higher than 1 M in A431 cells (Fig. 3A). To find out if the inhibition was in the transcriptional step or the translational step, cells were treated with 2 M GA for 9 h, and RNA was then extracted and examined by RT-PCR. This result indicated that the mRNA level of 12(S)-lipoxygenase and IKK␣ were reduced upon GA treatment in A431 cells, but no difference was observed in that of GAPDH with or without GA treatment (Fig. 3B). Furthermore, luciferase activity assay revealed that when the plasmid, pXLO7-1, containing the key promoter region of 12(S)-lipoxygenase localized at ϩ1 to Ϫ220 bp was transfected into A431 cells, the promoter activity was inhibited to 49, 42, and 28.5% upon treatment with 0.5, 1, and 2 M GA, respectively (Fig. 3C). Taken together, inhibition of Hsp90␣ by GA reduced the transcription of 12(S)-lipoxygenase in A431 cells.
Knockdown of Hsp90␣ Reduced the Expression of 12(S)-Lipoxygenase-GA can inhibit the activity of the Hsp90 family, including Hsp90␣, Hsp90␤, and Grp96 (32). To more specifically elucidate the role of Hsp90␣ in the transcription of 12(S)-lipoxygenase, siRNA was designed to knockdown the level of Hsp90␣ in cells. Plasmids, pSM2 and pSM2-shHsp90␣ were transfected into the AmphoPack-293 cells, respectively. After 72 h of incubation, viral particles of retrovirus-shHsp90␣ and retrovirus were collected in the supernatant. A431 cells were infected with the retrovirus or retrovirus-shHsp90␣ for 48 h, and the cell lysates were then collected for immunoblotting analysis. The results revealed that there was no significant change in Sp1 level, which served as a negative control, and both Hsp90␣ and 12(S)-lipoxygenase levels were reduced upon infection with the retrovirus-shHsp90␣ at an m.o.i. of 5 in A431 cells. Little change was observed when the cells were infected by the retrovirus (Fig. 4A). The levels of Hsp90␣ and 12(S)lipoxygenase were reduced to 24 and 50%, respectively after retrovirus-shHsp90␣ infection for 48 h, but there was no significant difference upon retrovirus infection (Fig. 4B). In addition, cells were infected with retrovirus at an m.o.i. of 10 or retrovirus-shHsp90␣ at an m.o.i. of 1, 5, and 10, and then co-transfected with 1 g of pXLO7-1. After 18 h of incubation, cell lysates were collected for the luciferase activity assay. These results revealed that the luciferase activity driven by the promoter of 12(S)-lipoxygenase was reduced to 72, 67, and 53% in cells treated with 0.5, 1, and 2 M GA, respectively, after normalization with that of retrovirus-only in A431 cells (Fig. 4C). Moreover, the RNA level of 12(S)-lipoxygenase decreased significantly after knockdown of Hsp90␣ by retrovirus-SM2-siHsp90␣, when no change was observed in GAPDH levels with or without GA treatment (Fig. 4D).

Effect of GA on Transcriptional Activity of 12(S)-Lipoxygenase Was Abolished with the Mutant 12(S)-Lipoxygenase Promoter at the Sp1
Binding Sites-To examine whether the effect of Hsp90␣ on the transcriptional activity of 12(S)-lipoxygenase was Sp1-dependent, three Sp1 binding sites (43), Sp1-1, Sp1-2, and Sp1-3, localized within Ϫ158 to Ϫ114 bp of the 12(S)-lipoxygenase promoter were mutated individually or together to study the relationship between Hsp90␣ and transcriptional activity of 12(S)-lipoxygenase (Fig. 5). The results show that the luciferase activity driven by the wild-type promoter of 12(S)-lipoxygenase was reduced to 26.5% (normalized with that driven by vector pXP-1 under GA treatment). In addition, the activity driven by the mutated 12(S)-lipoxygenase that single mutation on Sp1-1, Sp1-2, or Sp1-3 of the promoter were recovered to 60, 86, and 87.5%, respectively upon the FIGURE 1. Hsp90␣ can be recruited to the promoter of the 12(S)-lipoxygenase in cells. A, nuclear extracts were prepared from A431 cells and incubated with biotinylated oligonucleotide from the GC-rich region (Ϫ171 to Ϫ150 bp) of the 12(S)-lipoxygenase promoter. The samples were separated by 7% SDS-PAGE and immunoblotted for Sp1. B, samples from A were separated with an 8 -20% gradient two-dimensional SDS-PAGE, and then silver staining was performed. C, the spot shown with the arrow on the two-dimensional SDS-PAGE in B was collected and identified by MALDI-TOF MS. Sequences of five oligopeptides underlined were obtained, and those were completely identical with those of Hsp90␣. D, cells were fixed by 0.5% formaldehyde, and the genomic DNA was sonicated until the size was smaller than 500 bp. The samples were then analyzed by ChIP assay for Sp1 and Hsp90␣, and IgG was used as a negative control. The samples had 35 cycles of PCR and were separated on an agarose gel. treatment with GA. Interestingly, the level of activity was completely recovered when all three Sp1 binding sites were mutated with GA treatment. These results clearly indicated that Hsp90␣ was involved in the transcription of 12(S)-lipoxygenase in an Sp1-dependent fashion in A431 cells.
Nuclear Hsp90␣ Was Involved in the Transcription of 12(S)-Lipoxygenase-To study the importance of nuclear Hsp90␣ on the transcription of 12(S)-lipoxygenase, the DNA fragment of 12(S)-lipoxygenase localized within Ϫ220 to ϩ500 bp (T1, containing the promoter and partial coding region), Ϫ220 to ϩ1 bp (T2, containing the promoter only), and ϩ1 to ϩ500 bp (T3, containing the partial coding region) were used for the in vitro transcription assay (Fig. 6A). The DNA fragments were incubated with nuclear extracts in cells treated with or without GA in the presence of [ 32 P]UTP. RNA synthesis was subse-quently analyzed with 6% acrylamide gel containing urea (Fig. 6B). No signal was detected when either the promoter region (T2) or the coding region (T3) was used (Fig. 6B, lanes 1 and 2). However, a signal was observed at 500 bp in the absence or presence of RNase i, when the template containing both promoter and coding regions (T1) was used (Fig. 6B, lanes 3 and 4). When the nuclear extracts were treated with 1 or 5 M GA to block the Hsp90␣, the signals were decreased (Fig. 6B, lanes  5 and 6). The results revealed that RNA synthesis of 12(S)-lipoxygenase was inhibited under GA treatment.
Sp1 Binding to Promoter of 12(S)-Lipoxygenase Was Inhibited by GA-To study the mechanism of the Hsp90␣ effect on the target gene, A431 cells were treated with or without GA and fixed with formaldehyde. The nuclear extracts were prepared, and a ChIP assay was undertaken. These results revealed that Sp1 bound quite strongly to the pro-FIGURE 2. Hsp90␣ interacts with Sp1 in cells. A, nuclear extracts were extracted from A431 cells, and then incubated with anti-Sp1 and anti-IgG antibodies. The samples were analyzed with immunoblotting for Hsp90␣ and Sp1. B, GST-Sp1 was expressed in E. coli and purified by glutathione beads. 1 g of GST-Sp1 or GST was incubated with nuclear extracts of A431 cells at 4°C. The samples were separated with 7% SDS-PAGE and analyzed by immunoblotting using anti-Sp1 and anti-Hsp90␣ antibodies. The GST-beads were used as a negative control. C, GST-Sp1 and Hsp90␣ were expressed in E. coli. 1 g of GST-Sp1 and Hsp90␣ were used to do the in vitro direct interaction assay. D, cells were seeded and fixed, and the colocalization was then analyzed with anti-Hsp90␣, anti-tubulin ␣, and anti-Sp1 antibodies. After washing with PBS, samples were stained with FITC-anti-mouse antibodies for Hsp90␣ and tubulin ␣, and Cy5-anti-rabbit antibodies for Sp1. moter of 12(S)-lipoxygenase (Fig. 7A, lane 1). When the cells were treated with 0.5, 1, and 2 M GA, the Sp1 binding signals were decreased gradually (Fig. 7A, lanes 2-4). Consistent with the above results, the recruitment of Hsp90␣ was also decreased significantly upon GA in a dose-dependent manner (Fig. 7A, lanes 6 -9). In addition, no signal shown in lanes 5 and 10 of Fig. 7A was observed, indicating that the specific interaction between Sp1 and promoter was required. Furthermore, we also investigated the binding conditions of other transcription factors to the promoter of 12(S)-lipoxygenase with or without GA treatment by ChIP assay (Fig. 7B). The quantified data from several independent experiments indicated that in the presence of GA, recruitment of Sp1, Hsp90␣, TFIID (TBP), and acetyl-H3 to the promoter of 12(S)lipoxygenase was reduced to about 23, 65, 35, and 34%, respectively, but no obvious signal variation was detected when the actin or IgG was tested (Fig.  7C). These results demonstrated that Sp1 binding to the promoter of 12(S)lipoxygenase was influenced by the Hsp90␣ in A431 cells.  The samples were then analyzed by SDS-PAGE, and the expression of actin was used as an internal control. B, signals shown in A were quantified from three independent experiments. C, plasmid, pXLO7-1, was co-transfected into the retrovirusinfected A431 cells, and cells were then incubated for 1 day. The cell lysates were collected, and the transcriptional activity of 12(S)-lipoxygenase was analyzed by the luciferase assay. All of the experiments were performed three times at least independently, and statistical analysis was performed by Student's t test. D, both retrovirus-SM2-and retrovirus-SM2-siHsp90␣-infected A431 cells, and untreated cells were then incubated for 1 day. RNA was isolated, and the RNA levels of Hsp90␣, 12(S)lipoxygenase, and GAPDH were studied by RT-PCR.

DISCUSSION
Sp1 exerts its role in the regulation of target genes by binding to the GC-rich promoter region of the target gene. Numerous transcription factors cannot bind to DNA directly, but rather are recruited to the promoter through other transcription factors that directly bind to promoter. Therefore, Sp1 can be considered an important anchor protein that is able to recruit many factors to the promoter region of the target genes to carry out the transcription mechanism; such as recruiting c-Jun to enhance the transcriptional activity of the target gene(s) and recruiting p300 to promoter to acetylate the other transcription factors to regulate the transcription (44, 50 -52). In this study, we used the biotinylated GC-rich promoter region of 12(S)-lipoxygenase as a probe to detect the proteins recruited through Sp1 prior to two-dimensional SDS-PAGE analysis. Interestingly, after ruling out nonspecific binding by the mutated probe as shown in Fig. 1, many spots appeared in the two-dimensional gel. We identified the specific binding proteins by MALDI-TOF MS and observed that Hsp90␣ could be recruited to the promoter region of 12(S)-lipoxygenase through Sp1 in A431 cells. Other co-chaperones such as TCP-1, Cdc37, and Hsp70 were also found in the recruited complex (data not shown). Therefore, recruitment of the whole chaperone complex to the promoter of 12(S)-lipoxygenase occurred. Although Hsp90 mainly stays and functions in the cytosol, several reports showed that Hsp90 can translocate into the nucleus to regulate the nuclear retention of GR (53)(54)(55). In addition, several Hsp90␣ client proteins such as p53 involved in signal transduction pathway have been shown to enter the nucleus to execute their functions (56,57). In this study, two pieces of evidence supported that Hsp90␣ could be recruited to promoter of 12(S)-lipoxygenase in A431 cells. First, the results of DAPA and ChIP assay (Fig. 1) showed that Hsp90␣ could be recruited to the 12(S)-lipoxygenase promoter. Second, the immunoprecipitation and pull-down assays using the nuclear extracts (Fig. 2, A and B), and immunofluorescence microscopic analysis to study the co-localization between Sp1 and Hsp90 in vivo also revealed that Hsp90␣ could be recruited to the nucleus (Fig. 2D).
Next, the Hsp90␣ inhibitor, GA, which binds to the N terminus of A, the DNA fragments of 12(S)-lipoxygenase containing promoter and coding region (Ϫ220 to ϩ500 bp, T1), promoter (ϩ1 to Ϫ220 bp, T2), or coding region (ϩ1 to ϩ500 bp, T3) were prepared by PCR from A431 genomic DNA. B, the DNA fragments were then incubated with nuclear extracts prepared from cells with or without GA treatment for 1 h at 37°C containing [ 32 P]UTP. The synthesis of RNA in vitro was analyzed with 6% acrylamide gel. RNase i was used to prevent the degradation of synthesis RNA. Gels were then dried, and autoradiography was performed. All of the experiments were independently performed three times.

FIGURE 5. Hsp90␣ was involved in the Sp1-dependent transcription of 12(S)-lipoxygenase in cells.
The plasmids: pXP-1,which was an empty vector, pXLO7-1, which contained Ϫ220 to ϩ1 bp region of 12(S)-lipoxygenase, mSp1-1, mSp1-2, and mSp1-3 in which the Ϫ167 GGC Ϫ165 was changed to Ϫ167 TTT Ϫ165 , the Ϫ156 GG Ϫ155 to Ϫ156 TT Ϫ155 , and the Ϫ120 GGCGGG Ϫ115 to Ϫ120 TTATTT Ϫ115 , respectively, mSp1-4 in which all the three Sp1 binding sites, were mutated transfected into A431 cells, and cells were then treated with 1 M GA for 18 h. Cell lysates were collected, and the luciferase assay was performed. The relative percentage of inhibition by GA treatment was obtained by normalizing the values of luciferase activity driven by the promoter of 12(S)-lipoxygenase and various mutants, with or without GA treatment. All of the experiments were independently performed three times.
Hsp90␣ was used to block the activity of the Hsp90␣, and siRNA of Hsp90␣ was further used to reduce the level of Hsp90␣. Both experiments revealed that the expression of 12(S)-lipoxygenase was inhibited in the presence of inactive Hsp90 or Hsp90 knockdown conditions (Figs. 3 and 4). However, previous studies indicated that Hsp90 interacts with P52 (rIPK) to inhibit the P58 (IPK) to mediate downstream control of PKR activity and eIF2␣ phosphorylation. The dephosphorylated eIF2␣ cannot inhibit the translational mechanism (58), and therefore, Hsp90␣ enhances translation efficiency. In our present study, we still could not rule out the possibility that the translational activity of the 12(S)-lipoxygenase was affected under GA treatment or Hsp90␣ siRNA treatment in A431 cells. However, our results show that the transcriptional activity was affected under GA treatment in A431 cells. First, the results shown in Fig. 3 revealed that the mRNA level was reduced under GA treatment. Second, the reporter assay indicated that the promoter activity of 12(S)-lipoxygenase was reduced to about 25% under GA treatment. Third, the in vitro transcription assay shown in Fig. 6 revealed that the RNA synthesis level was reduced under GA treatment.
According to the previous studies, Hsp90 can affect many proteins in cell cycle and signal transduction (59 -62). Indeed, in our experiments, there is no doubt that Hsp90␣, recruited by Sp1 to the nucleus, was important for the transcription of 12(S)-lipoxygenase in A431 cells. The reporter assay shown in Fig. 5 revealed that mutation at three Sp1 binding sites localized on the promoter of 12(S)-lipoxygenase resulted in reduction of the inhibition level because of GA treatment in A431 cells.
These results revealed that interaction between Sp1 and Hsp90␣ was very important for the transcription of 12(S)-lipoxygenase. Recently, many kinases have been shown to translocate to the nucleus with their substrate to carry out their functional biological roles within the nuclei (63). Hsp90␣ has large conformational flexibility to form a multitude of dynamic co-chaperone complexes, which contribute to functional diversity (64). This assists a wide range of the client proteins, to stabilize After fixation with 0.5% formaldehyde, nuclear extract was extracted and incubated with anti-Sp1 antibodies and anti-Hsp90␣. The samples were then analyzed by Southern blotting. B, the other binding factors, TFIID, histone 3, Hsp90␣, and actin were also studied by using anti-TFIID (TBP), antiacetyl-H3, and anti-Hsp90 antibodies. The samples were then analyzed by Southern blotting. C, the signals shown in B from several independent experiments were quantified, and the actin and IgG were used as negative controls. The letter n represents the number of independent experiments. their structures. A large number of the client proteins have a consensus sequence TPR (65)(66)(67). A remarkable proportion of its substrates are proteins related to the protein kinases, cell cycle control, and signal transduction (68,69). There is no TPR sequence inside the Sp1, and hence no direct interaction between Sp1 and Hsp90␣ occurred. Rather, this recruitment by Sp1 is indirect. The role of Hsp90␣ recruited to Sp1 might affect other factor(s) recruited by Sp1 to regulate the transcriptional activity of 12(S)-lipoxygenase. In summary, in this study, we demonstrated that Hsp90␣ could be recruited by Sp1 to the promoter of 12(S)-lipoxygenase to modulate its transcriptional activity, by affecting the binding of Sp1 to the gene promoter in A431 cells.