| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
B/CCAAT-Binding Transcription Factor-1 Complex
Department of Veterinary Physiology & Pharmacology (C.Q., T.N., J.S., I.S., R.S.) and Department of Veterinary Anatomy and Public Health (R.B.), Texas A&M University, College Station, Texas 77843-4466
Address all correspondence and requests for reprints to: Stephen H. Safe, Department of Veterinary Physiology & Pharmacology, Texas A&M University, 4466 TAMU, College Station, Texas 77843-4466. E-mail: ssafe{at}cvm.tamu.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
B (NFB) proteins are required for hormone responsiveness. The p65 subunit of NFB was identified in both nuclear and cytosolic fractions of untreated MCF-7 cells; however, formation of the nuclear NFB complex was E2 independent. Hormonal activation of constructs containing p53 promoter inserts (-106 to -40) and the GAL4-p65 fusion proteins was inhibited by the intracellular Ca2+ ion chelator EGTA-AM and Ca2+/calmodulin-dependent protein kinase (CaMK) inhibitor KN-93. Constitutively active CaMKIV but not CaMKI activated p65, and treatment of MCF-7 cells with E2 induced phosphorylation of CaMKIV but not CaMKI. The results indicate that hormonal activation of p53 though nongenomic pathways was CaMKIV-dependent and involved cooperative p65-CTF-1 interactions. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
p53 is expressed in estrogen receptor 
(ER)- positive T47D breast cancer cells and levels are significantly decreased in cells maintained in charcoal-stripped serum or serum-free media (18, 19, 20). 17ß-Estradiol (E2) has minimal effects on p53 levels in cells grown in serum-containing media, whereas antiestrogens such ICI 164,384, tamoxifen or 4'-hydroxytamoxifen decrease p53 protein. However, in T47D cells grown in media containing charcoal-stripped serum, E2 significantly increased levels of p53 protein (18). It was also reported that E2 induced chloramphenicol acetyltransferase (CAT) activity in T47D or MCF-7 breast cancer cells transfected with the PICAT construct that contains a 2.4-kb insert from the p53 gene promoter linked to a CAT reporter gene (19). In the present study, we have investigated the mechanism of E2-mediated transcriptional activation of p53 in MCF-7 human breast cancer cells. Treatment with E2 increased p53 mRNA (
2-fold) and protein (2- to 3-fold) levels and luciferase (luc) reporter gene activity in MCF-7 cells transfected with p53–1 (containing the -1800 to +12 region of the p53 gene promoter). Subsequent deletion and mutational analysis of the p53 gene promoter has identified a minimal 67-bp region (-106 to -40) that is required for hormonal activation of p53-derived constructs and the key cis-acting elements include CTF-1 and nuclear factor B (NFB) binding sites where CTF-1 acts as a DNA-bound cofactor for NFB through direct interactions with p65. E2 activates p53 through nongenomic pathways and studies with kinase inhibitors show that p65 is activated in breast cancer cells through E2-dependent activation of calmodulin kinase IV (CaMKIV), and this is consistent with a recent report showing that CaMKIV also activates p65 in CV-1 and HeLa cells (21).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
also show that a time-dependent increase in p53 mRNA levels was observed after treatment with E2 and a maximum 2-fold increase was noted after 20 h. Ten- and 100-nM E2 alone significantly induced (1.7- to 2.2-fold) luc activity in MCF-7 cells transfected with p53-1, which contains a p53 gene promoter insert from -1800 to +12 (Fig. 1B). Moreover, after cotransfection with ER, there was an enhanced (7.8-fold) induction of reporter gene activity after treatment with E2. Enhanced gene expression after cotransfection with ER has previously been reported using other E2-responsive constructs (22, 23, 24, 25, 26), and ER was cotransfected in subsequent experiments in this study.
|
). Deletion of the -1800 to -1199 region increased basal activity (p53-1/p53-2) and further deletion to -592 (p53-3) resulted in a 4-fold decrease in basal activity in MCF-7 cells. Basal activity increased 2-fold in cells transfected with p53-4, suggesting inhibitory elements within the -1800 to -1199 and -592 to -312 region of the p53 gene promoter. Marked loss of luc activity was observed in cells transfected with p53-7, p53-8 and p53-9 containing the -66 to +12, -44 to +12, and -106 to -67 p53 gene promoter inserts. E2 responsiveness was observed in cells transfected with p53-1 to p53-6 (3- to 7.8-fold) indicating that the -106 to +12 region of the promoter was the minimal sequence required for hormone activation. Moreover, a 3'-deletion to give p53-9 (-106 to -67) resulted in loss of E2 responsiveness.
|
B, E2F, Sp1-like proteins, and an E-box (Fig. 2B
). Mutation of the E2F binding site (p53-6 m) or deletion of the -30 to +12 region of the promoter (p53-10) did not result in loss of E2 responsiveness, suggesting that the E2F site is not required for E2 action. Deletion of the 5' CTF-1/YY1 (p53-11 and p53-12) and mutation or deletion of the NFB (p53-10 m2 and p53-14) binding sites resulted in loss of E2-induced transactivation, whereas mutation or deletion of the overlapping GC-rich/E-box motifs (p53-10 m1 and p53-13) did not affect this response. These results suggest that hormone-induced transcriptional activation of constructs derived from the p53 gene promoter require the -106 to -40 region containing intact CTF-1/YY1 and NFB binding sites.
3) Protein Interactions with the p53 Gene Promoter
Protein interactions with the E2-responsive region of the p53 gene promoter were determined using nuclear extracts from MCF-7 cells and [32P]-labeled oligonucleotides from -106 to -30 (Fig. 3). Preliminary gel mobility shift studies indicated that ER did not directly bind this region of the p53 promoter. The upstream fragment (from -106 to -67) has been reported to bind both CTF-1 and YY1 (27). Results in Fig. 3B show that nuclear extracts bind [32P]-106/-67 to form one major retarded band (lane 1) and intensity of this band is decreased only after competition with 50-fold excess of unlabeled -106/-67 and consensus CTF-1 oligonucleotides (lanes 2 and 4), but not by consensus oligonucleotides that bind NF-Y, CP2, C/EBP, NFB, Sp1 (lanes 3, 5–7, and 9) or an estrogen-responsive element (ERE) (lane 8). ER does not interact with this region (-106 to -67) of the promoter and oligonucleotide competition experiments suggest that CTF-1 binds this sequence. In a separate experiment, it was shown that [32P]-labeled YY1 probe formed a band with MCF-7 cell nuclear extracts; however, the unlabeled YY1 oligonucleotide did not competitively decrease binding to [32P]-106/-67 (data not shown).
|
B (lane 4), and Sp1 (GC-rich) (lane 5) oligonucleotides decreased intensity of the retarded bands (Fig. 3C). The higher molecular weight band A was preferentially decreased by coincubation with the consensus Sp1 oligonucleotide, whereas intensities of the lower molecular weight bands B and C were preferentially decreased by the consensus NFB oligonucleotide. The lowest molecular weight band D was affected (decreased intensity) only after competition with consensus NF-Y oligonucleotide (lane 3), whereas competition with an ERE did not affect intensities of bands A–D. Using slightly altered binding conditions that favor NF-Y but not Sp1 binding (band A) (Fig. 3D), only bands B, C, and D were formed, and results of unlabeled oligonucleotide competition experiments were similar to those observed in Fig. 3C. The failure of an ERE to decrease band intensities indicates that ER also does not bind directly to the -66 to -30 region of the p53 gene promoter.
Confirmation of specific protein interactions with [32P]-66/-30 were also determined in antibody supershift experiments (Fig. 3E) using the two incubation conditions described in Fig. 3, C and D. In this experiment where bands A, B, C, and D were observed (lane 1, Fig. 3E), supershift experiments with Sp1 and Sp3 antibodies (lanes 3 and 4) gave at least two supershifted bands, and NF-YA antibodies also gave a supershifted complex (lane 5). Using incubation conditions that eliminated Sp1-DNA complex formation (Fig. 3F) (lane 1), antibodies to p65 and p50 proteins supershifted bands B and C, respectively (lanes 3 and 4). Results from antibody supershift experiments confirm oligonucleotide competition studies showing that NF-Y, Sp1, Sp3, p65, and p50 bind to the overlapping NF-Y/NFB/GC-rich sites in the -66 to -30 region of the p53 gene promoter.
4) Physical and Functional Interactions of NFB and CTF-1
Interactions of CTF-1 and p65 with the p53 gene promoter were also investigated using a chromatin immunoprecipitation assay in which cells were treated with formaldehyde to form DNA-protein cross-links. After sonication, and immunoprecipitation by p65 or CTF-1 antibodies, PCR was used to determine binding of p65 and CTF-1 protein to the -196 to -6 region of the p53 gene promoter (Fig. 4A). The results indicated that both p65 and CTF-1 were bound to the promoter in untreated cells and in cells treated with E2 for 15 min; the slight decrease in transcript levels observed after 15 min may be due to several factors including decreased antigen recognition due to recruitment of other promoter binding proteins after treatment with E2. cAMP response element binding protein (CREB) antibodies were used as a negative control for the p53 promoter immunoprecipitation and transcripts were not detected. In a separate experiment, we used a modified chromatin immunoprecipitation (ChIP) assay method to investigate interactions of CaMKIV, ER, and Sp1 interactions with the proximal region of the p53 gene promoter (Fig. 4B). This modified assay used an increased number of cells to isolate immunoprecipitable chromatin complexes, and PCR products were separated by agarose gel electrophoresis and detected by ethidium bromide staining as described (28). CaMKIV was cross-linked to the p53 gene promoter, and both ER and Sp1 also bound to this region of the promoter. Detection of Sp1 binding is consistent with the GC-rich motif in the promoter. In addition, we have previously detected ER associated with GC-rich promoters in the presence or absence of E2 (Refs. 29, 30 , and our unpublished results), and this is consistent with the reported ligand-independent direct interactions between ER and Sp1 proteins (31, 32). In this study, ER only weakly interacted with the p53 gene promoter in untreated cells; however, enhanced interactions were observed after treatment with E2 for up to 120 min. As an additional control experiment, we also observed ER interactions with the cathepsin D gene promoter (-294 to -54) after 0, 15, and 30 min (Fig. 4C), whereas Sp1 protein was bound after 30 min (positive control). This results is consistent with the role of ER and Sp1 proteins in activation of cathepsin D (25), and previous studies (33) have also reported the ER antibodies immunoprecipitate this region of the cathepsin D gene promoter. In contrast, ER and Sp1 antibodies did not immunoprecipitate a region of exon 2 of the cathepsin D gene (negative control) (Fig. 4D).
|
B and CTF-1 motifs on the p53 gene promoter are required for estrogen action (Figs. 1 and 2), we examined the interactions of these proteins in immunoprecipitation (Fig. 5A) and a mammalian two-hybrid assay (Fig. 5B). In coimmunoprecipitation assays (Fig. 5A), [35S]p65 protein (lane 1) was immunoprecipitated by p65 antibody (lane 3), whereas this same antibody did not immunoprecipitate [35S]CTF-1 (lane 5). However, after incubation of [35S]CTF-1 with p65 protein, p65 antibody immunoprecipitated [35S]CTF-1 (lane 4), demonstrating physical interactions between p65 and CTF-1 proteins. Whole cell lysates from E2-treated MCF-7 cells were immunoprecipitated with CTF-1 antibody (lane 8) or with IgG as a control (lane 7). Lane 8 was whole cell lysate alone (Fig. 5A). All samples were analyzed by Western blot analysis with p65 antibody, which showed positive immunostaining in lanes 7 and 8. Cellular interactions between p65 and CTF-1 were investigated in a mammalian two-hybrid system (Fig. 5B) in MCF-7 cells. p65 or a C-terminal region of p65 (p65c1) were fused to the DNA binding domain (DBD) of the yeast GAL4 protein in the pM vector. CTF-1 was fused to the acidic VP16 protein in the pV16 vector, and the cells were also transfected with a construct containing 5 tandem GAL4 response elements (pGAL4-luc). Transfection with the GAL4 fusion proteins containing p65 (pM-p65) or p65c1 (pM-p65c1) increased luc activity by 256- and 483-fold, respectively. Cotransfection with pV-CTF-1 decreased activities of the fusion proteins by 65–71%, thus indicating that non-DNA-bound CTF-1 squelched pM-p65 and pM-p65c1-mediated activities. In contrast, pV-CTF-1 did not affect the activity of pM3VP16, indicating that the interaction between CTF-1 and p65 was specific. Thus, CTF-1 appears to act as a cofactor for p65 in the context of the p53 gene promoter.
|
B and NFB interactions with other DNA-bound transcription factors such as Sp1 or AP1 (29, 30, 31, 32, 34, 35, 36, 37) or through nongenomic induction of kinase pathways (38, 39, 40, 41, 42, 43, 44). Transcriptional activation of p53–10 by E2 in MCF-7 cells was not affected by cotransfection with dominant negative (dn) NF-Y and Sp1 expression plasmids (Fig. 6A), and the latter result is consistent with the E2 responsiveness of p53–10 m1, which is mutated in the GC-rich Sp1 binding site. These results suggest that, although ER and Sp1 can interact with this region of the p53 gene promoter (Fig. 4B), ER/Sp1 interaction with the GC-rich site is not required for hormone-induced transactivation. In contrast, activation of p53–10 by E2 was significantly inhibited after cotransfection with the IB super repressor (IB-SR), which cannot be phosphorylated and degraded (45) and after cotransfection with precursor-derived inhibitor (pdI), which contains the C-terminal region of the p105 precursor of p50 protein component of the NFB heterodimer. Purified pdI can dissociate preformed complexes of bacterial p50 bound to DNA and also inhibit purified NFB (46). In addition, the proteasome inhibitor MG132 also blocked transcriptional activation of p53–10 by E2. The results obtained for MG132, pdI and IB-SR are consistent with an important role for the NFB site in transcriptional activation of the p53 gene promoter by E2, and this was supported by the hormone nonresponsiveness of p53–10 m2 (Fig. 2) containing an NFB site mutation. The importance of the NFB pathway in hormonal activation of p53 was further confirmed in studies showing that E2 induced a 2- to 3-fold increase in p53 protein levels after treatment for 2 or 6 h, and the induction response was blocked in cells transfected with IB-SR (Fig. 6B).
|
B, it is possible that E2 either induces formation of nuclear NFB or directly activates NFB (p65/p50) through nongenomic pathways that involve induction of kinase activities (38, 39, 40, 41, 42). The responsiveness of MCF-7 cells to other activators of NFB was investigated using IL-1ß, which caused a significant decrease in cytosolic IB protein after treatment for 15 or 30 min (but not 60 min). Recovery of protein levels after 60 min was blocked by cycloheximide (CHX) (Fig. 6C, lane 9). In contrast, E2 had no effect on levels of cytosolic IB or nuclear p65 protein (Fig. 6C, lanes 6 and 7), whereas IL-1ß increased nuclear p65 protein levels after treatment for 15, 30, or 60 min (Fig. 6C, lanes 3–5). NFB binding to the -60 to -30 region of the p53 gene promoter was analyzed by gel mobility shift assay (Fig. 6D), and the results showed that the IL-1ß alone (lanes 2–5) or in combination with cycloheximide (lane 6) increased binding, whereas cycloheximide alone (lane 7) or E2 alone (lanes 8 and 9) did not induce NFB binding. Binding specificity was demonstrated by competition with unlabeled consensus NFB oligonucleotide and antibody supershifts (lanes 11 and 12); in addition, unlabeled NF-Y oligonucleotide decreased intensity of the NF-Y-DNA complex band. Although IL-1ß significantly increased the formation of the DNA-protein complex, treatment with this cytokine did not induce luc activity in MCF-7 cells transfected with p53–13 and did not affect the E2-induced response (Fig. 6E). These data suggest that treatment of MCF-7 cells with E2 did not markedly affect nuclear NFB levels of DNA binding.
Initial studies in MCF-7 cells showed that IB protein was detected in nuclear and cytosolic extracts and was unaffected by treatment with E2 (data not shown). Application of fluorescence imaging techniques using p65 antibodies showed significant nuclear staining in untreated cells, and this was only slightly enhanced after treatment with E2 (Fig. 6F). These results are consistent with the results of transactivation, Western blot and gel mobility shift assays indicating that E2 does not activate p53 by inducing nuclear accumulation of NFB.
6) E2 Directly Activates NFB through the C-Terminal Activation Domain of p65
The direct activation of p65 and CTF-1 by E2 through nongenomic pathways were investigated using GAL4 fusion proteins and a pGAL4-luc construct in MCF-7 cells. The results in Fig. 7A show that E2 directly activates p65 C-terminal transactivation domain, but not CTF-1 in one-hybrid system indicating that p65 is the major downstream target of E2. p65c2 contains the C-terminal 30 amino acids from p65 fused to the GAL4-DBD; this construct was cotransfected with pGAL4-luc and ER, and treatment with E2 induced 3.8-fold increase in luc activity. This response was not blocked by cotransfection with pdI indicating that E2 directly activated the p65 C-terminal transactivation domain that does not interact with pdI. E2 did not induce luc activity through pM-CTF-1 or pM-CTF-1c (a fusion of GAL4-DBD with CTF-1 C-terminal 100 amino acids), and CTF-1 or CTF-1c showed less transactivation potential than NFB under these conditions in MCF-7 cells.
|
, B and C, show that 3 µM KN-93 does not block hormonal activation of pM-65c2 or p53–10, whereas 30 µM KN-93 and the intracellular calcium ion chelator EGTA-AM blocked induction in cells transfected with the same constructs. A recent report indicated that 30 µM KN-93 inhibited Ca2+/CaM-dependent protein kinase-mediated activation of p65 in CV-1 and HeLa cells. Moreover, both constitutively active CaMKI and CaMKIV (but not CaMKII) also activated p65, and CaMKIV directly associated with p65 in vivo and in vitro (21). In contrast, our results show that constitutively active CaMKIVc but not CaMKIc activated pM-65c2 in MCF-7 cells (Fig. 7D). Furthermore, E2 also significantly induced phosphorylation of CaMKIV after treatment for 30 min (Fig. 7E). However, in parallel studies, increased phosphorylation of CaMKI was not observed even though CaMKI was detected by Western blot analysis (data not shown). These results show that hormonal activation of p65/NFB is dependent on E2-induced release of intracellular Ca2+ and subsequent phosphorylation of CaMKIV. These observations are consistent with previous reports showing that E2 activated intracellular Ca2+ in MCF-7 cells (40) and CaMKIV activated p65 in CV-1 and HeLa cells (21). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Deletion of the CTF-1 site did not decrease basal activity of p53-12 (-80 to -30), whereas interaction of CTF-1 (or YY1) with this site was important for basal activity of the p53 promoter in HeLa cells (27). Previous studies have reported that ER/Sp1 and NFYA interactions with GC-rich and CCAAT sites, respectively, were necessary for induction of E2F1 in MCF-7 cells (34); however, these interactions were not required for hormone activation of p53-10 (Fig. 2). Although gel mobility shift analysis (Fig. 3) and chromatin immunoprecipitation assays (Fig. 4) confirmed that multiple proteins in MCF-7 cell nuclear extracts bound the -106 to -30 region of p53 gene promoter, results of transactivation studies (Fig. 2) suggest that the NFB and CTF-1 sites are required for E2 responsiveness and protein binding to the NFB site is also critical for basal activity (Fig. 2).
The CTF-1 binding site in the minimal promoter of p53 (-106 to -40) is conserved in human, rat, and mouse p53 promoters. Occupancy of this site varies in a tissue-specific manner; it is bound primarily by YY1 in nuclear extracts of rat testis and spleen and by CTF-1 in extracts of liver and prostate, and this may facilitate tissue-specific control of p53 gene expression. Both YY1 and CTF-1 function as activators when bound to this site in HeLa cells (27). However, in this study, only CTF-1 bound this element (Fig. 3) and in a chromatin immunoprecipitation assay with CTF-1 antibodies (Fig. 4), there was direct evidence for CTF-1 (and NFB) binding to the proximal region of the p53 gene promoter. Previous studies have shown that ER can activate a chimeric protein containing the proline-rich C-terminal domain of CTF-1 fused to the GAL4 DBD in transient transfection studies in HeLa cells; however, these results were obtained using a promoter-reporter construct containing a ERE insert (47). In MCF-7 cells, we did not observe hormonal activation of CTF-1 using pGAL4-luc and either pM-CTF-1 (full coding sequence) or pM-CTF-1c (fusion with C-terminal 100 amino acids). However, when fused to pVP16, the CTF-1 fusion protein significantly squelched activity of pM-p65 (Fig. 5), indicating that CTF-1 interacts in vivo with p65. Immunoprecipitation data using whole cell lysates and in vitro expressed p65 and CTF-1 also confirmed physical interaction of these proteins (Fig. 5A), suggesting that DNA-bound CTF-1 acts as cofactor of NFB for activation of p53. In contrast, the NMDA receptor agonist quinolinic acid also induced p53 in neuronal cells through NFB; however, this response was dependent on increased nuclear uptake of p65 and did not require CTF-1 as a cofactor (48).
A major pathway for activation of NFB signaling involves phosphorylation of IB followed by dissociation of the NFB-IB complex and rapid proteasome-dependent degradation of IB (reviewed in Refs. 49 49 and 50). This is also accompanied by the rapid nuclear translocation of NFB and formation of a nuclear NFB complex. This was also observed after treatment of MCF-7 cells with IL-1ß (Fig. 6, B and C) and previously described for this cytokine in other cell lines (49, 50). Activation of p53 constructs and p53 protein by E2 requires dissociation of IB and proteasome-dependent degradation (Fig. 6, A and B), but increased binding of nuclear NFB complex to DNA was not observed after treatment with E2. In contrast, IL-1ß enhanced NFB binding to the p53 promoter, but did not induce luc activity or enhance E2-induced luc activity in cells transfected with p53–10. These results indicate that E2 does not enhance formation of nuclear NFB. Immunohistochemistry studies (Fig. 6F) indicate that p65 is nuclear in approximately 40% of untreated MCF-7 cells, and this percentage is only slightly increased (approximately 60%) after treatment with E2. This suggests that E2 may directly activate NFB in MCF-7 cells.
Other studies have demonstrated that treatment-induced dissociation of cytosolic IB from NFB is not sufficient for activation through NFB site; moreover, degradation of IB and NFB-DNA binding are not always correlated with activation of target genes (51, 52, 53, 54). For example, Bergmann et al. (51) showed that activation of an NFB-dependent reporter gene by TNF and IL-1ß in A549 human alveolar cells is blocked by phospholipase C and protein kinase C inhibitors, which did not affect IB degradation or nuclear translocation/DNA binding of NFB. SB203580, a p38 MAPK inhibitor, completely inhibited TNF-induced synthesis of IL-6 and expression of a reporter gene containing a minimal promoter with two NFB elements; however, neither TNF-induced DNA binding of NFB nor TNF-induced phosphorylation of its subunits was modulated by SB203580 (52, 55), and other studies gave similar results (53, 56).
Several reports show that activation of kinase pathways are required for induction of NFB-dependent activities, and these kinases include NIK/MAP3K (57), MAPK P38 and ERK (52, 55), MEKK1 (58), phosphatidylcholine-specific phospholipase C (51, 59) and protein kinase C (51, 60, 61, 62), tyrosine kinase (53), PKA (63), casein kinase II (CK II) (64), Akt/PKB (65, 66), and CaMKIV (21). The role of kinase activation by E2 through nongenomic pathways was investigated using a series of kinase inhibitors, and only inhibitors of Ca2+-dependent kinases (EGTA-AM and KN-93) block hormonal activation of pM-65c2 and p53–10 (Fig. 7, B and C). A recent study reported that E2 increases mobilization of intracellular Ca2+ in MCF-7 cells and this response has been linked to activation of the MAPK pathway (40). Results of inhibitory studies with the intracellular Ca2+ chelator EGTA-AM indicate that Ca2+ mobilization is also important for activation of p65 in MCF-7 cells (Fig. 7, B and C); however, this response was not blocked by the MAPKK inhibitor PD98059. Intracellular Ca2+ also plays an important role in activation of Ca2+/CaM-dependent kinases (67), and a recent study showed that these kinases stimulate activation and phosphorylation of p65 in CV-1 and HeLa cells (21). It was reported that constitutively active CaMKI and CaMKIV (but not CaMKII) activated p65 and this response was inhibited by the Ca2+/CaM-dependent kinase inhibitor KN-93 (21). In contrast, our results show that constitutively active CaMKIV but not CaMKI activated p65 in MCF-7 cells (Fig. 7D). Hormonal activation of both CaMK1 and CaMKIV was further investigated in MCF-7 cells, treated with [32P]-labeled orthophosphate, and the results showed that E2 induced phosphorylation of CaMKIV (Fig. 7E), whereas phosphorylation of CaMKI in MCF-7 cells was not detected (data not shown). These data suggest that hormonal activation of p65/NFB in MCF-7 cells is CaMKIV-dependent and related to increased mobilization of Ca2+ in this cell line (40). ChIP has identified nuclear ER bound to the E2-responsive region of the p53 gene promoter (Fig. 4B), and ER can interact with p65, CTF-1 and Sp1 proteins (31, 32, 47, 68), which also bind elements within this region. Results of transactivation, gel mobility shift and kinase inhibitor assays suggest that p53 is activated by E2 through nongenomic pathways. However, the detection of ER bound to the p53 promoter suggests that nuclear ER may contribute to the observed responses and this is currently being investigated.
In summary, our results show that activation of p53 gene expression in MCF-7 cells by E2 is dependent on nongenomic activation of CaMKIV, which results in activation of p65. This pathway is consistent with previous reports showing that E2 activates release of intracellular calcium (40, 69, 70, 71) and CaMKIV stimulates NFB through phosphorylation of p65 (21). Although E2 stimulates proliferation of MCF-7 cells, the concomitant induction of p53, a protein associated with growth inhibition, is not unexpected. For example, estrogen also induces proteasome-dependent degradation of ER protein (72, 73, 74, 75) and expression of genes such as early growth response 1 (76) that can act as a negative transcription factor. p53 and BRCA1 are genes associated with inhibition of cell growth, DNA repair and apoptosis, and both BRCA1 and p53 are induced by E2 in breast cancer cells (18, 19, 20, 77, 78). Interestingly, both proteins physically interact with ER and inhibit ER-mediated transactivation (79, 80), suggesting that hormonal activation of p53 and BRCA1 may contribute to regulatory pathways that lead to inhibition of E2-induced cell growth and differentiation. Other mitogens and growth regulatory proteins also induce inhibitory pathways that temporally regulate the magnitude and duration of their responses (81, 82). The role of p53 as negative regulator of hormone-induced proliferation of MCF-7 cells and the mechanisms of E2-dependent nongenomic activation of calcium-dependent kinase pathways are being further investigated in this laboratory in ER-positive and ER-negative cancer cell lines.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
B p65 (sc-109X), p50 (sc-1190X), IB- (sc-371), CaMKIV, CaMKI, and preimmune IgG were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and NF-YA antibody from Rockland Inc. (Gilbertsville, PA). Poly deoxy-(insolinic-cytidylic) acid, restriction enzymes, and T4-polynucleotide kinase, and pGL2 luc reporter vector, rabbit reticulocyte lysate system were purchased from Promega Corp. (Madison, WI) and Roche Molecular Biochemicals (Indianapolis, IN). Reporter Lysis Buffer and Luciferase Reagent for Luciferase studies were purchased from Promega Corp. ß-Galactosidase (ß-gal) Reagent was purchased from Tropix (Bedford, MA). Plasmid preparation kit was purchased from QIAGEN (Santa Clarita, CA). The mammalian MATCH MAKER two-hybrid assay kit was purchased from CLONTECH Laboratories, Inc. (Palo Alto, CA). All other chemicals and biochemicals were the highest quality available from commercial sources. Lab-Tek Chamber slides was purchased from Nalge Nunc International (Naperville, IL). InstantImager and LumiCount were purchased from Packard (Meriden, CT).
Oligonucleotides and Plasmids
The p53 cDNA probe was purchased from Geneka Biotechnology Inc. (Montréal, Québec, Canada). p53 promoter- derived oligonucleotides, primers employed in plasmid construction, and consensus oligonucleotides for EMSA were synthesized by Genosys/Sigma (The Woodlands, TX) or by the laboratory of Dr. James Derr (Department of Veterinary Pathobiology, TAMU). Consensus oligonucleotides used in EMSA competition experiments are Sp1 (5'-AGC TTA TTC GAT CGG GGC GGG GCG AGC g-3'), ERE (5'-GTC CAA AGT CAG GTC ACA GTG ACC TGA TCA AAG TT-3'), NF-Y (AGA CCG TAC GTG ATT GGT TAA TCT CTT), CTF-1 (TTT TGG ATT GAA GCC AAT ATG ATA A), LSF (GAG CAA GCA CAA ACC AGC CAA), NFB (AGT TGA GGG GAC TTT CCC AGG C). All the sequences listed in this paper are sense strands. They might have been synthesized in both sense and anti-sense strands, annealed and phosphorylated when necessary. Some sequences may contain additional restriction sites at flanking sites for cloning, however the linkers are not indicated.
The p1 CAT construct containing the human p1 p53 promoter was kindly provided by Dr. David Reisman (University of South Carolina, Columbia, SC), and the p53 promoter fragments were PCR amplified with appropriate primers. The p53 promoter luc reporters were constructed using pGL2 (Promega Corp.) for cloning of p53–1 to -8, and TpGL2 for p53–9 to -14. The TpGL2 vector was constructed by inserting a minimal TATA sequence (GCT GTA GGG TAT ATA ATG GAT CA) with linker into the BglII and the HindIII sites in pGL2. The reverse primers used for constructing critical mutated reporter p53–10 m2 was AGT CTC GAG CAC ATG GGA GGG Gag cTC aaC ggT CCC ATC AAC C. All ligation products were transformed into TOP10F' competent cells (Invitrogen, Carlsbad, CA). Plasmids were confirmed by restriction enzyme mapping and DNA sequencing. High-quality plasmids for transfection were prepared using QIAGEN Plasmid Megaprep Kit (QIAGEN). The human ER expression plasmid was kindly provided by Dr. Ming-jer Tsai (Baylor College of Medicine, Houston, TX). The dnNF-YA (Dr. Michael Schweizer, Institute of Food Research, Norwich Research Park, Colney, Norwich, UK), pPacP65 and pdI (Drs. Claus Scheidereit and Vigo Heissmeyer, Max-Delbruck-Centrum fur Molekulare Medizin, Berlin, Germany), IB-SR (Drs. Albert Baldwin and Jacqueline Norris, University of North Carolina, Chapel Hill, NC), pCTF-1 (Dr. Robert Tjian, University of California, Berkeley, CA) and dnSp1 (Dr. Gerald Thiel, University of the Saarland Medical School, Hamburg, Germany) plasmids were generous gifts. Constitutively active CaMKIV and CaMKI were kindly provided by Dr. Eric Olson (University of Texas Southwestern Medical School, Dallas, TX) and Dr. Anthony Means (Duke University Medical School, Durham, NC), respectively. The Gal4 reporter containing 5X Gal4 DBD (pGAL4-luc) and p65c2 (previously named Gal4-p65
Sma) containing p65 C-terminal 30 amino acids were kindly provided by Dr. Marty Mayo (University of North Carolina, Chapel Hill, NC). The pM-p65 was constructed by inserting p65 cDNA sequence from pPacP65 into EcoRI and XbaI sites of pM (CLONTECH Laboratories, Inc.), and p65c1 is an EcoRI restriction of pM-p65. The pV-CTF-1 was constructed by inserting CTF-1 cDNA sequence from pCTF-1 into EcoRI and XbaI sites of pVP16 (CLONTECH Laboratories, Inc.).
Northern Blot Analysis
MCF-7 cells were seeded in 5% CSS DME/F12 media for 24 h, and in serum-free DME/F12 media for another 48 h. Cells were treated in fresh serum-free media and harvested at various times. Total RNA was isolated using STAT-6 Kit. Twenty micrograms of total RNA were diluted in 2x RNA sampling buffer (20% formaldehyde; 1.65% Na2HPO4, pH 6.8; 63.5% formamide; 1x loading buffer), and separated on 1.2% agarose gel with 1 M formaldehyde in 1x running buffer (20 mM Na2HPO4, 2 mM CDTA, pH 6.8). After transfer onto Hybond-N nylon membrane, and prehybridized and hybridized in NENhybe solution [750 mM NaCl, 50 mM NaH2PO4, 5 mM EDTA (pH 7.4); 10% dextran sulfate, 0.1% polyvinyl pyrolidine, 0.1% Ficoll, 0.1% BSA, 1% sodium dodecyl sulfate (SDS)] at 65 C without/with 32P-ATP-labeled 40-oligomer p53 cDNA probe for 16 h. The membrane was washed 15 min with 1x washing buffer containing 150 mM NaCl, 10 mM NaH2PO4, 1 mM EDTA (pH 7.4), and 2% SDS. Blots were visualized by autoradiography and bands were quantitated on InstantImager (Packard). The membrane was then stripped (0.1x saline sodium citrate, 1% SDS) and rehybridized with ß-tubulin probe as a control.
Transient Transfection and Luciferase Activity Assay
MCF-7 cells were seeded in 5% CSS DME/F12 media in 60-mm plate 1 d before transfection using calcium phosphate-DNA coprecipitation method. Reporter plasmid DNA (2.5 µg) and 1.0 µg ß-gal DNA, with or without 1.0 µg wild-type human ER DNA, were used for transfection in promoter analysis; 0.25–2.0 µg of dnNF-Y, pdI, IB-SR, and dnSp1 were used for cotransfection, and their related empty vectors were added to balance total transfected DNA amounts; 2.5 µg GAL4-luc reporter, 0.25 µg pM or pVP16 and their related fusions were used in mammalian two-hybrid system transfection. After incubation for 16–20 h, cells were washed with PBS and treated with E2 or DMSO (as control) for 24 h if applicable in fresh media. Cells were then lysed with 400 µl of 1x reporter lysis buffer. Thirty microliters of cell extract were subjected to luc and ß-gal assays. LumiCount was used to quantitate luc and ß-gal activities, and the luc activities were normalized to ß-gal activity.
Preparation of Cytosolic and Nuclear Extracts
MCF-7 cells were seeded in 5% CSS DME/F12 media in 150-mm plates for 2 d, and treated as indicated. Cells were then harvested in 3 ml of PBS. After spinning down, cells were resuspended in HED buffer (25 mM HEPES; 1.5 mM EDTA; 1.0 mM dithiothreitol, pH 7.6). After incubation on ice for 10 min, cells was pelleted by centrifuging at 2000 x g for 10 min. Supernatant was discarded, cells were lysed in HEGD buffer (25 mM HEPES; 1.5 mM EDTA; 1.0 mM dithiothreitol; 10% glycerol, pH 7.6), syringed with 21-gauge needles 6x, checked with trypan blue for cell breakage. Cytosolic protein (supernatant) was collected after centrifuging at 4000 x g for 15 min. Nuclei were then washed with HEGD buffer (5x), and lysed with HEGDK buffer (25 mM HEPES; 1.5 mM EDTA; 1.0 mM dithiothreitol; 10% glycerol; 500 mM KCl, pH 7.6) on ice for 60 min. Nuclear protein was collected after centrifuging at 12,000 x g for 30 min at 4 C, aliquoted and stored at -70 C.
Western Blot Analysis
MCF-7 cells were seeded in 5% CSS DME/F12 media for 2 d, and pretreated with CHX and treated with E2 or IL-1ß as indicated. Cytosolic and nuclear protein were collected, and protein samples were heated at 100 C for 5 min, separated on 10% SDS-PAGE at 160V for 3 h in 1x running buffer (25 mM Tris-HCl; 192 mM glycine; 0.1% SDS, pH 8.3), and transferred to polyvinylidenedifluoride membrane (Amersham Pharmacia Biotech) at 100 V for 2 h at 4 C in 1x transfer buffer (48 mM Tris-HCl, 39 mM glycine, 0.075% SDS). The polyvinylidenedifluoride membrane was blocked in 5% milk-TBS (10 mM Tris-HCl, 150 mM NaCl, pH8.0) with gentle shaking for 1 h, and incubated in fresh 5% milk-TBS with 1:3000 (for p53 and CaMKIV), 1:1000 (for IB) or 1:4000 (for p65) primary antibody (Santa Cruz Biotechnology, Inc.) for 1 h with gentle shaking. After washing in 1x TBS for 5 min (3x), 1:4,000 secondary antibody in 5% milk-TBS was incubated with shaking for 1 h. The membrane was then washed again in TBS buffer for 1 h, incubated in 10 ml of chemiluminescent substrate (electrochemiluminescence) (NEN Life Science Products) for 1.0 min, and exposed to Kodak (Rochester, NY) X-Omat AR autoradiography film immediately. Band intensities were evaluated by scanning laser densitometry (Sharp Electronics Corp.) and background subtraction using Zero-D Scanalytics software (Scanalytics Corp.). The membrane was reprobed with Sp1 antibody as control.
Immunocytochemistry
MCF-7 cells were seeded in 5% CSS DME/F12 media to 50% confluency on Lab-Tek Chamber slides (Nalge Nunc International), treated with E2 or DMSO for various time and washed with 20 mM PBS. Slides were fixed for 10 min in -20 C methanol, air-dried, and washed in 0.3% Tween/PBS (Use 0.3% Tween in 20 mM PBS for all subsequent washing steps) for 5 min. The slides were then blocked with isotype IgG for 1 h. The NFB p65 antibody (Santa Cruz Biotechnology, Inc.) was diluted in antibody dilution buffer (1% BSA in 0.3% Tween/PBS), and 1:750 p65 antibody was added to the well. Slides were incubated overnight at 4 C in a humid chamber to prevent evaporation of the antibody solution. After washing slides in 0.3% Tween/PBS for 10 min (3x), diluted second antibody (raised against the same species as the primary antibody) was added. The secondary antibody was directly conjugated with fluorochrome. Slides were incubated for 2 h at room temperature in a dark humid chamber, and then washed in 0.3% Tween/PBS for 10 min (4x), and rinsed in deionized water. Slides were mounted in glycerol/phenylenediamine and visualized. Samples without primary antibody or secondary antibody were also used as control.
Gel EMSA
Nuclear extracts were prepared from MCF-7 cells treated with 10 nM of E2 for 2 h utilizing cells maintained in MEM media supplemented with 10% FBS. Synthetic oligonucleotides were annealed, and labeled with T4-polynucleotide kinase and [
-32P]-ATP. EMSA was performed by incubating 3–5 µg of nuclear extract in 25 µl of binding buffer (25 mM HEPES; 1.5 mM EDTA; 1.0 mM dithiothreitol; 2 mM MgCl2; 10% glycerol; 100 mM KCl, pH 7.6) in the presence of 1 µg deoxy-(insolinic-cytidylic) at 4 C for 5 min. 32P-labeled oligonucleotide (50,000 cpm) was added and incubated for an additional 15 min at 25 C. Excess unlabeled DNA used for competition and antibodies used for supershifts were added before 32P-labeled oligonucleotides. For analysis of Sp1-DNA binding, 0.1 mM ZnCl2 was included in the incubation reaction. The reaction mixture was separated on 5% polyacrylamide gel (acrylamide:bisacrylamide, 30:0.8) at 140 V in 1x TBE buffer (0.09 M Tris-HCl; 0.09 M boric acid; and 2 mM EDTA, pH 8.3). The gel was dried and exposed to PhosphoScreen and visualized on STORM.
In Vivo Phosphorylation
MCF-7 cells were seeded in 100-mm dishes with 10% FBS MEM media at 30% confluency for 24 h, and cultured in 5% CSS DME/F12 media for another 24 h. Cells were then changed to serum-free DME/F12 and incubated with 200 µCi 32P-orthophosphate for 3 h. Cells were then treated with 10 nM E2 for the indicated times, and washed with PBS (3x). Whole cell lysate (WCL) was collected and immunoprecipitated with CaMKIV antibody. Protein bound beads were suspended in 1x Western sampling buffer, and heated at 100 C for 5 min. Protein samples were separated on 10% SDS-PAGE. The gel was dried and exposed to PhosphoScreen and quantitated on STORM.
Coimmunoprecipitation/In Vitro
For immunoprecipitation by antibodies against NFB p65, in vitro transcribed/translated proteins of CTF-1 and NFB were obtained using the rabbit reticulocyte lysate system (Promega Corp.). Protein was labeled using [35S]methionine (Amersham Pharmacia Biotech). Equal volumes of 10 µl of CTF-1 and NFB proteins were incubated for 1 h at room temperature. To this mixture, NFB p65 antibodies were added and samples were subsequently incubated for 2 h on ice. Protein G-agarose beads (Santa Cruz Biotechnology, Inc.) were added and incubated at 4 C for 3–5 h. Immunoprecipitation of protein G-bound proteins was conducted according to manufacturer’s recommendations and subsequently separated by 7% SDS-PAGE electrophoresis and visualized by autoradiography.
Coimmunoprecipitation/Western Blot
MCF-7 cells were seeded in 5% CSS DME/F12 media in 150-mm plates for 2 d, and treated as indicated. Cells were rinsed with PBS at room temperature, and harvested in 0.5 ml of HEGD buffer [1x PBS, 1% Nonidet P-40 (NP-40), 0.5% sodium deoxycholate, 0.1% SDS, 100 µg/ml phenylmethylsulfonyl fluoride (PMSF)]. Cells were lysed by freeze-thawing in liquid nitrogen/ice water (3x) and brought to 0.5 M in sodium chloride by addition of 4 M sodium chloride, 20 mM HEPES (pH 7.4), then transferred using a syringe with a 21- gauge needle to a fresh tube, and incubated with 10 µl of 10 mg/ml PMSF for 30–60 min on ice. Supernatant was collected as WCL after microcentrifuging at 10,000 x g for 10 min at 4 C. WCL was precleared by adding 0.25 µg of control IgG with 20 µl protein G-agarose conjugate to approximately 1 ml WCL, and incubating at 4 C for 30 min with shaking. Beads were pelleted by centrifuging at 1000 x g for 5 min at 4 C, and supernatant (cell lysate) was collected. To 200 µg of the supernatant, 2 µg of CTF-1 or normal IgG antibody was added. The reaction mixture was incubated at 4 C for 1 h, and mixed with 20 µl of resuspended protein A/G-agarose beads and incubated for 1 h on a rocker platform at 4 C. Pellet was collected by contrifugation at 1000 x g for 5 min at 4 C, and washed with HEGDK buffer (4x). After the final wash, the pellet was resuspended in 40 µl of 1x electrophoresis buffer for Western blot analysis using p65 antibodies as described above.
ChIP Assay
MCF-7 breast cancer cells were grown in 150-mm tissue culture plates to 80–95% confluency and treated with 10 nM E2 for 15 min. Formaldehyde was then added to the medium to give a 1% solution and incubated with shaking for 10 min at 20 C. After addition of glycine (0.125 M) and incubation for 10 min, the media was removed, cells were washed with PBS and 1 mM PMSF, scraped, and collected by centrifugation. Cells were then resuspended in swell buffer (85 mM KCl, 0.5% NP-40, 1 mM PMSF, 5 µg/ml leupeptin and aprotinin at pH 8.0) and homogenized. Nuclei were isolated by centrifugation at 1500 x g for 30 sec, then resuspended in sonication buffer (1% SDS; 10 mM EDTA; 50 mM Tris, pH 8.1), and sonicated for 45–60 sec to obtain chromatin with appropriate fragment lengths (500–1000 bp). This extract was then centrifuged at 15,000 x g for 10 min at 0 C, aliquoted and stored at -70 C until used. The cross-linked chromatin preparations were diluted in buffer (1% Triton X; 100 mM NaCl; 0.5% SDS; 5 mM EDTA; and Tris, pH 8.1), and 20 µl of Ultralink protein A or G or A/G beads (Pierce Chemical Co., Rockford, IL) was added per 100 µl chromatin and incubated for 4 h at 4 C. A 100-µl aliquot was saved and used as the 100% input control. Salmon sperm DNA, specific antibodies, and 20 µl Ultralink beads were added and the mixture incubated for 6 h at 4 C. Samples were then centrifuged; beads were resuspended in dialysis buffer, vortexed for 5 min at 20 C, and centrifuged at 15,000 x g for 10 sec. Beads were then resuspended in immunoprecipitation buffer (11 mM Tris; 500 mM LiCl; 1% NP-40; 1% deoxycholic acid, pH 8.0) and vortexed for 5 min at 20 C. The procedures with the dialysis and immunoprecipitation buffers were repeated (3–4 times), and beads were then resuspended in elution buffer (50 nM NaHCO3, 1% SDS, 1.5 µg/m sonicated salmon sperm DNA), vortexed, incubated at 65 C for 15 min. Supernatants were then isolated by centrifugation and incubated at 65 C for 6 h to reverse protein-DNA cross-links. Wizard PCR kits (Promega Corp.) were used for additional DNA cleanup and PCR using [32P]-deoxy-CTP was used to detect the presence of promoter regions immunoprecipitated with commercially available CREB1, p65, CTF-1, ER, or Sp1 antibodies (Santa Cruz Biotechnology, Inc.). An improved ChIP assay using an increased amount of cross-linked chromatin for the immunoprecipitation assay was carried out as described (28). The same p53-specific primers were used in a two-step PCR reaction (95–60 C). PCR products were run on a 2% agarose gel and stained with ethidium bromide. The following primers were used for PCR analysis of immunoprecipitated promoter regions:
p53: Fw (-196), 5'-GCG GTA CCC CAG GTC GGC GAG AAT CC-3'; Rv (-6), 5'-GGG CTC GAG TCT AGA CTT TTG AGA AGC-3'.
Cathepsin D: Fw (-294), 5'-TCC AGA CAT CCT CTC TGG AA-3'; Rv (-54, 5'-GGA GCG GAG GGT CCA TTC-3'.
Cathepsin D (exon 2): Fw (+2469), 5'-TGC ACA AGT TCA CGT CCA TC-3'; Rv (+2615), 5'-TGT AGT TCT TGA GCA CCT CG-3'.
Statistical Analysis
Statistical differences between different groups were determined by ANOVA and Scheffé’s test for significance. The data for all transient transfection studies are presented as mean ± SD for at least three replicate determinations for each treatment group and were observed in at least two sets of experiments.
|
FOOTNOTES |
|---|
Abbreviations: CaMKIV, Calmodulin kinase IV; CAT, chloramphenicol acetyltransferase; ChIP,Chromatin immunoprecipitation; CHX, cycloheximide; CREB, cAMP response element binding protein; CSS, charcoal-stripped FBS; CTF-1, CCAAT-binding transcription factor-1; DBD, DNA binding domain; DMSO, dimethylsulfoxide; dn, dominant negative; E2, 17ß-estradiol; ER, estrogen receptor; ERE, estrogen-responsive element; FBS, fetal bovine serum; ß-gal, ß-galactosidase; IB-SR, IB super repressor; luc, luciferase; NFB, nuclear factor B; NF-Y, nuclear factor-Y; NP-40, Nonidet P-40; pdI, precursor-derived inhibitor; PMSF, phenylmethylsulfonylfluoride; SDS, sodium dodecyl sulfate; WCL, whole cell lysate.
Received for publication January 7, 2002. Accepted for publication May 10, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
) or 10 nM E2 (
), and luc activity was determined as described in Materials and Methods. Luciferase activity was significantly (P < 0.05) activated in the DMSO group (*) and induced by E2 (**). Effects of kinase inhibitors and EGTA on hormonal activation of pM-p65c2 (B) and p53-10 (C). Cells were transfected with pM-p65c2 or p53-10 and treated with E2, DMSO alone or in the presence of different kinase inhibitors or EGTA, and luc activity was determined as described in Materials and Methods. Significant (P < 0.05) induction by E2 is indicated by an asterisk. Results are presented as means ± SD for at least three replicate experiments for each treatment group. D, Activation of pM-p65c2 by constitutively active CaMKI and CaMKIV. MCF-7 cells were transfected with pM-p65c2 alone and in combination with CaMKIc or CaMKIVc (constitutively active) (0.1 or 0.5 µg) and luc activity determined as described in Materials and Methods. Significant (P < 0.05) induction is indicated (*), and results are express as mean ± SD for three replicate determinations for each treatment group. E, Phosphorylation of CaMKIV. MCF-7 cells were pretreated with [32P]orthophosphate for 3 h and then treated with DMSO (C) or 10 nM E2 for 2, 30 or 240 min; whole cell lysates were obtained, immunoprecipitated (IP) with CaMKIV antibody, and analyzed by SDS-PAGE/Western blot (WB) analysis and phosphoimaging as described in Materials and Methods. Increased phosphorylation of CaMKIV after treatment with E2 for 30 min was observed in duplicate experiments. In parallel studies using CaMKI antibodies, phosphorylation of CaMKI was not observed.