Stimulation of p300-mediated transcription by the kinase MEKK1.

p300 and CREB-binding protein (CBP) are related transcriptional coactivators that possess histone acetyltransferase activity. Inactivation of p300/CBP is part of the mechanism by which adenovirus E1A induces oncogenic transformation of cells. Recently, the importance of p300/CBP has been demonstrated directly in several organisms including mouse, Drosophila, and Caenorhabditis elegans where p300/CBP play an indispensable role in differentiation, in patterning, and in cell fate determination and proliferation during development. CBP/p300s are modified by phosphorylation during F9 cell differentiation as well as adenovirus infection, suggesting that phosphorylation may play a role in the regulation of p300/CBP activity. Here we show that the mitogen-activated/extracellular response kinase kinase 1 (MEKK1) enhances p300-mediated transcription. We identify several domains within p300 that can respond to MEKK1-induced transcriptional activation. Interestingly, activation of p300-mediated transcription by MEKK1 does not appear to require the downstream kinase JNK and may involve either a direct phosphorylation of p300 by MEKK1 or by other non-JNK MEKK1-directed downstream kinases. Finally, we present evidence that p300 is important for MEKK1 to induce apoptosis. Taken together, these results identify MEKK1 as a kinase that is likely to be involved in the regulation of the transactivation potential of p300 and support a role of p300 in MEKK1-induced apoptosis.

p300 was initially identified as an adenovirus E1A-associated cellular protein in coimmunoprecipitation experiments (1,2). The ability of adenovirus E1A to immortalize cells and to induce full morphological transformation is dependent on its ability to interfere with the functions of p300 and its related protein CBP 1 (3)(4)(5) and members of the retinoblastoma family members (6 -9). In addition, inactivation of p300 and CBP is necessary for E1A to inhibit differentiation (10 -12). These experiments suggest that p300/CBP play an important role in cell proliferation and differentiation. Recent experiments in mouse, Drosophila, and Caenorhabditis elegans have provided direct evidence that p300/CBP are crucial in both differentiation and cell proliferation during development (13)(14)(15). p300 and CBP are both transcriptional cofactors that can acetylate histones and transcription factors due to their histone acetyltransferase (HAT) activity (16 -19). This suggests that p300 and CBP may regulate transcription in part by modifying the chromatin structure as well as activities of transcription factors (18 -20). Consistent with this, a C. elegans homolog, CBP-1, has been shown to promote endoderm differentiation by antagonizing the histone deacetylase activity, highlighting the importance of the biological significance of the HAT activity of CBP-1 (13). Additional mechanisms that account for p300/CBPmediated transcriptional activation involve direct interactions with the basal factors (4,21,22) as well as with the RNA polymerase II complex (23).
Although a great deal is known about p300/CBP as transcriptional regulators, very little information is available about mechanisms that regulate their transcriptional activity. Several lines of evidence suggest that phosphorylation may regulate p300/CBP activity. First, p300/CBP have been shown to be phosphoproteins whose phosphorylation state alters during F9 cell differentiation and in adenovirus infection (24). Second, kinases such as protein kinase A and mitogen-activated kinases have been shown to activate CBP-mediated transcription (25,26). Third, p300 has been reported to be phosphorylated in vitro by cyclin-dependent kinases such as cdc2 and cdk2 (27). In this report, we provide evidence that the mitogen-activated protein kinase/ERK kinase kinase 1 (MEKK1) stimulates p300mediated transcription in vivo in a manner that is independent of its downstream kinase JNK.
MEKK1 is a 196-kDa protein which, when cleaved by caspases in response to genotoxic agents, is capable of inducing apoptosis (28,29). MEKK1 propagates the signal by activating its downstream kinase, MKK4, which in turn activates JNK (30,31). Activated JNK translocates into the nucleus where it modifies the activity of transcription factors such as AP-1 (32,33). MEKK1 has also been shown to activate NFBand c-Mycmediated transcription, the former involving the activation of the IB␣ kinase (34). We find that the activated MEKK1 can significantly increase GAL4-p300-mediated transcription. Consistently, nocodazole, which is an extracellular stimulus of MEKK1 (35,36), can also stimulate p300-dependent transcription. Interestingly, this activation appears to be independent of the downstream kinase JNK since a dominant negative JNK fails to abrogate MEKK1-induced, p300-mediated transcription. In addition, mutations of all the possible JNK sites within subdomains of p300 responsive to MEKK1 had no effect on their response to MEKK1. Although the mechanisms by which MEKK1 induces p300-mediated transcription are still unclear, one possible scenario is a direct phosphorylation of p300 by MEKK1. Consistent with this possibility, we find that MEKK1 can robustly phosphorylate p300 directly in vitro. Significantly, we also find that the activated form of MEKK1 is present exclusively in the nucleus, overlapping with the endogenous p300. In addition, we find the transfected inactive MEKK1 in the cytoplasm, and it appears to localize some of the p300 proteins to the punctate MEKK1 signals. Finally, our results also show that MEKK1-induced apoptosis is impaired by the inhibition of p300 expression. Taken together, our results suggest that phosphorylation of p300 induced by MEKK1 modifies its transcriptional activity, which may be an important event in the apoptosis induced by MEKK1.
Luciferase Assay-HeLa cells were transfected using Fugene 6 transfection reagent (Roche Molecular Biochemicals) with indicated reporters and cDNA expression vectors. After 40 h, luciferase activity was determined as described (38). The luciferase activities were normalized on the basis of ␤-galactosidase activity of cotransfected Rous sarcoma virus ␤-galactosidase vector.
Assay for the Effect of Nocodazole on p300-dependent Transcription-U2OS cells were transfected with mammalian expression plasmids encoding either GAL4 DNA-binding domain (G4, 1 g) alone or G4-p300 (1 g) together with the target reporter G4-luciferase (0.5 g). One hour before harvest, the cells were treated with two different doses (0.5 and 1.0 g/ml) of nocodazole. Cell extracts were prepared and used for luciferase assay described above.
Expression and Purification of Recombinant Proteins-GST-p300 fusion proteins were induced and purified as described (34). His 6 MEKK1⌬ was purified from baculovirus-infected Sf9 cell lysates using nickel-nitrilotriacetic acid-agarose as described previously (34).
In Vitro Kinase and Apoptosis Assays-About 5 g of purified GST fusion proteins were suspended in 30 l of kinase buffer (20 mM MOPS, pH 7.2, 2 mM EDTA, 10 mM MgCl 2 , 0.1% Triton X-100) and incubated with either 0.5 g of His 6 MEKK1⌬ or 1 g of recombinant JNK (Stratagene) and 10 Ci of [␥-32 P]ATP at 37°C for 30 min. The kinase reactions were terminated by the addition of Laemmli sample buffer. Proteins were resolved using 12% SDS-PAGE gels and dried for autoradiography.
Nocodazole stimulation of MEKK1 kinase activity was measured by the activation of JNK. Briefly, proliferating U20S cells were treated with either Me 2 SO solvent control or nocodazole at the indicated concentrations. Cell lysates were prepared 1 h post-treatment, and the activation of MEKK1 was analyzed by Western blotting with an anti-JNK1 antibody (SC-474, Santa Cruz Biotechnology) as described previously (39).
Apoptosis assays were performed using MCF-7 cells that express either an active p300 or inactive p300 ribozymes as described previously (40). Activated MEKK1 was cotransfected into MCF-7 cells along with green fluorescent protein (GFP) cDNA. Apoptotic cells were scored on the basis of GFP and Hoecht stain 48 h after transfection.
Cell Fixation and Immunofluorescence Microscopy-U2OS cells were transiently transfected with full-length HA-MEKK1 or FLAG-MEKK1⌬ cDNA. Thirty six hours post-transfection, the cells were fixed in 3.7% paraformaldehyde at 25°C for 10 min. After 5 min of washing in PBS, the fixed cells were permeabilized in 0.2% Triton X-100 (25°C for 15 min) followed by 5 min of washing in PBS. The cells were then incubated with primary antibody for 1 h at 25°C. Anti-HA epitope tag monoclonal antibody (Babco) and anti-FLAG epitope tag monoclonal antibody (Sigma) were used at a 1:200 dilution; anti-p300 polyclonal antibody C-20 (Santa Cruz Biotechnology) was diluted 1:100. After three 10-min PBS washes, coverslips were incubated in secondary antibody for 1 h at 25°C. Fluorescein isothiocyanate-conjugated goat anti-mouse (Jackson ImmunoResearch) and Cy3 Red-conjugated goat anti-rabbit (Jackson ImmunoResearch) secondary antibody were used at a dilution of 1:400. For the DAPI staining, the coverslips were also incubated in 1ϫ PBS with 1:10,000 dilution of DAPI for 10 min. Vectashield 1200 media was used for slide mounting (Vector Labs). Epifluorescence microscopy was conducted using a Nikon microscope. Images were acquired by a Sony camera and analyzed using Adobe Photoshop version 5.0.

MEKK1⌬
Activates GAL4-p300-mediated Transcription-To determine whether MEKK1 affects p300-mediated transcription, a constitutively activated MEKK1 mutant, MEKK1⌬ (34), was cotransfected into HeLa cells along with GAL4-p300 and a GAL4 luciferase reporter construct. Fig. 1A shows that MEKK1⌬ augmented GAL4-p300-mediated transcription about 20-fold compared with vector control. In contrast, a catalytically inactive mutant of MEKK1, MEKK1⌬ K432M (lysine 432 converted to methionine), had little effect on GAL4-p300mediated transcription (Fig. 1B). This indicates that stimulation of p300-mediated transcription by MEKK1 is dependent on the kinase activity of MEKK1. Furthermore, MEKK1⌬ did not affect transcription mediated by GAL4-Sp1Q2 (Fig. 1A), a previously characterized glutamine-rich activation domain of Sp1, suggesting that the ability of MEKK1⌬ to up-regulate p300mediated transcription was not through a general enhancement of the basal transcription machinery. The ability of MEKK1⌬ to activate GAL4-p300-mediated transcription was not restricted to HeLa cells as similar results were obtained in COS-7 cells (data not shown). These findings suggest that a kinase pathway involving MEKK1 can regulate GAL4-p300mediated transcription.
Nocodazole Stimulates p300-dependent Transcription-To determine whether extracellular stimuli known to activate MEKK1 can stimulate p300-dependent transcription, we examined the ability of nocodazole (35,36) to regulate GAL4-p300-mediated transcription. As shown in Fig. 2A, MEKK1 kinase activity is stimulated by nocodazole in a dose-dependent manner as shown by its ability to induce JNK phosphorylation (compare lanes 2 and 3 with lane 1). When the kinase activity of MEKK1 is stimulated, we find consistent enhancement of GAL4-p300-mediated transcription by nocodazole in a dose-dependent manner, although at a modest level (Fig. 2B, lanes 5 and 6). Taken together, activated MEKK1 provided either exogenously (transfection) or endogenously (conversion of an inactive MEKK1 to an active one by nocodazole) can stimulate GAL4-p300-mediated transcription.
Identification of Domains of p300 Involved in Its Response to MEKK1-Previously, we and others (21,22) have shown that at least two independent activation domains are present in the N-and C-terminal portions of p300. We wished to determine whether the transcriptional activity of these two p300 domains could be regulated by MEKK1⌬. Fig. 3 shows that the transcriptional activity of the N-terminal (aa 1-596) and the Cterminal regions (aa 1257-2414), when fused to GAL4, were highly responsive to MEKK1⌬. In contrast, the middle region of p300 (aa 744 -1571), which itself has no detectable transcriptional activity (22), did not respond at all to MEKK1⌬.
We next sought to identify a minimal domain within the N-terminal and C-terminal regions of p300 that were responsible for the transcriptional activation by MEKK1⌬. As shown in Fig. 3, the response of the large N-terminal region of p300 to MEKK1⌬ was mediated by two subdomains, aa 2-337 (N1, 32-fold) and aa 302-667 (N2, 89-fold) The N2 subdomain contains the C/H1 domain (aa 347-411) and is three times more responsive than N1. Further deletions that remove aa 302-406 (inclusive of the C/H1 domain) and aa 567-667 from N2 did not significantly reduce its response to MEKK1, suggesting that p300 aa 407-566 (N3) is sufficient to mediate part of the re-sponse of p300 to MEKK1⌬. Taken together, these results suggest that there are at least two transactivation domains within the N terminus that are responsive to MEKK1⌬, p300 aa 2-337 and p300 aa 407-566, and that the C/H1 domain is not involved in the N-terminal response of p300 to MEKK1⌬.
The C-terminal deletion mutants of p300 were also analyzed for the same purpose. Fig. 3 shows that MEKK1⌬ enhanced the transcriptional activity of GAL4-p300 (aa 1737-2414) (C1) 32fold. In contrast, MEKK1⌬ only weakly activated the transcriptional activity of GAL4-p300 (aa 1945-2414) (C2). These results suggest that p300 aa 1737-1945 is an important domain within the C-terminal half of p300 for responding to MEKK1⌬. Amino acids 1737-1945 of p300 contain the carboxyl portion of the C/H3 domain (aa 1653-1817) that is rich in cysteines and histidines (12). The C/H3 domain has been shown to be an important surface that mediates a number of important protein-protein interactions (12,21,(41)(42)(43). The above data suggest that the C-terminal domain of p300 that mediates the response to MEKK1 may reside somewhere between aa 1737 and 1945. We therefore tested a GAL4-p300 construct that includes this region (GAL4-p300 aa 1709 -1913) (C3). As shown in Fig. 3, MEKK1⌬ activates transcription mediated by GAL4-p300 C3 (aa 1709 -1913) by 20-fold. This is a significant transcriptional induction, although slightly less than the fold of activation (32-fold) observed for GAL4-p300 C1 (aa 1737-2414) in response to MEKK1⌬. Thus, a region including part of the C/H3 domain of p300 is involved in the transcriptional re- sponse of the C-terminal region of p300 to MEKK1⌬.
MEKK1 Stimulation of GAL4-p300-mediated Transcription Does Not Require JNK-It has been well documented that MEKK1 activates the JNK family members of kinases (30,31,44). To determine whether JNK1 was involved in the activation of GAL4-p300-mediated transcription induced by MEKK1, we first asked whether JNK is necessary for the MEKK1 response mediated by the subdomains of p300 described above. We tested a dominant negative inhibitor of JNK1 (JNK1 APF) (32) for its ability to block MEKK1⌬-mediated activation of GAL4-p300-mediated transcription. As shown in Fig. 4, JNK1 (APF) was unable to block MEKK1⌬ induction of transcription by either GAL4-p300N1 (aa 2-337), GAL4-p300N2 (aa 302-667), or GAL4-p300C1 (aa 1737-2414). In contrast, the dominant negative JNK1 mutant APF potently inhibited GAL4-cJunmediated transcription 8-fold as expected (Fig. 4). These finding therefore suggest that JNK is not necessary for MEKK1 to enhance p300-mediated transcription. MEKK1 has also been shown to activate the IB␣ kinases (34). By using a dominant negative form of the IB␣ kinase (S32A/S36A), we found no evidence that this kinase is involved in mediating the ability of MEKK1⌬ to enhance p300-mediated transcription (data not shown).
The above results predict that these subdomains of p300 are either not substrates for JNK1, or the JNK phosphorylation sites residing in these subdomains are not important for MEKK1-induced, p300-mediated transcription. To determine whether JNK1 can phosphorylate the MEKK1-responsive p300 subdomains, we performed in vitro kinase assays using the various p300 domains fused to GST as substrates for recombinant JNK. Equal amounts of GST proteins were used for the in vitro kinase assays (data not shown). Fig. 5 shows that recombinant JNK phosphorylates GST-p300 aa 1709 -1913 (C3) very strongly (lane 4). In contrast, recombinant JNK barely phosphorylated GST-p300 aa 2-337 or GST-p300 aa 302-667 (Fig.  5, lanes 2 and 3). Similar results were found in JNK immune complex kinase assays using the same GST-p300 substrates (data not shown). These results indicate that only the C3 domain of p300 was a potential substrate for JNK, whereas the N-terminal domains were not. The fact that the N-terminal domains of p300 are poor JNK substrates is consistent with the above finding (Fig. 5) that JNK is not involved in MEKK1-induced transcription mediated by the N-terminal domains of p300.
The finding that JNK phosphorylates GST-p300 aa 1709 -1913, which can also mediate a MEKK1 response, appears to be contradictory to the observation that the p300 C-terminal domain-mediated MEKK1 response is unaffected by a dominant negative form of JNK (Fig. 4). To resolve this issue, we determined whether JNK phosphorylation of this domain is correlated with its ability to mediate a MEKK1 response. We therefore mutated all potential JNK sites with the consensus sequence serine/proline or threonine/proline within this domain. Because of the large number of proline-directed serines and threonines within this region, we made a series of cluster mutations designated A-D shown in Fig. 6A. Equal amounts of the mutant GST fusion proteins (Fig. 6C) were then tested for their ability to be phosphorylated by recombinant JNK. Fig. 6B shows that compared with wild-type p300 aa 1709 -1913, mutation of proline-directed serines and threonines to alanines within either clusters A or B partially abolished the ability of recombinant JNK to phosphorylate GST-p300 aa 1709 -1913 (lanes 2 and 3). Mutation of serines and threonines to alanines within clusters C or D did not inhibit the ability of JNK to phosphorylate GST-p300 aa 1709 -1913 (data not shown). However, mutant GST-p300 aa 1709 -1913, which combines the mutations in both clusters A and B (mut A/B), totally lost the ability to respond to JNK-induced phosphorylation (Fig. 6B, lane 4).
We next analyzed the ability of this JNK phosphorylationdefective mutant for its ability to mediate an MEKK1 response. As shown in Fig. 6D, mutation of all potential JNK sites (clusters A-D) did not inhibit the response of this GAL4-p300 aa 1709 -1913 mutant to MEKK1⌬ compared with its wild-type counterpart. Although the N-terminal subdomain of p300, N2 (aa 302-667), is not phosphorylated by JNK1 in vitro (Fig. 5), we nevertheless mutated all possible JNK sites in this domain and analyzed the effect the mutations. As expected, mutation of the potential JNK sites in the N2 subdomain had little effect on its ability to mediate MEKK1-induced transcriptional activation (Fig. 6D). Taken together, these results support the notion that the downstream kinase of MEKK1, JNK1, is not essential for MEKK1-stimulated transcriptional activation mediated by p300. MEKK1 Can Phosphorylate the N Terminus of p300 in Vitro-Given the fact that JNK1 is not involved in the response of p300 to MEKK1, nor are the other potential downstream kinases such as p38 and IB␣, we considered the possibility that MEKK1⌬ may be able to directly phosphorylate p300. Baculovirus MEKK1⌬ was purified as described previously (34) and used in kinase assays with GST-p300 fragments as substrates. Fig. 7A shows that baculovirus MEKK1⌬ strongly phosphorylated GST-p300 aa 302-667 (lane 3) as well as GST-p300 aa 300 -528 (data not shown). As a control, MEKK1⌬ did not phosphorylate GST alone as shown in lane 1 (Fig. 7A). The ability of MEKK1 to phosphorylate GST-p300 aa 302-667 was also observed in immunocomplex kinase assays with MEKK1 (data not shown). In contrast, GST-p300 aa 2-337 (N1) (lane 2) and GST-p300 aa 1709 -1913 (C3) (lane 4) were weakly phosphorylated by MEKK1⌬ (compare lanes 2 and 4 with lane 3). The phosphorylation of GST-p300 aa 302-667 as opposed to GST-p300 aa 2-337 and GST-p300 aa 1709 -1913 was not due to differences in protein amount as Fig. 7B shows relatively equal levels of protein by Coomassie Blue staining. These results are consistent with the possibility that stimulation of GAL4-p300 aa 302-667-mediated transcription by MEKK1⌬ may occur via direct phosphorylation of p300 aa 302-667 by MEKK1. Furthermore, the inability of MEKK1 to phosphorylate GST-p300 N1 and GST-p300 C3 (Fig. 7A, lanes 2 and 4) suggests that a yet unidentified MEKK1-directed kinase other than JNK may be involved in activating GAL4-p300-mediated transcription.

Analyses of Subcellular Localization of MEKK1 and p300 -
The model that MEKK1 may regulate p300 activity by directly phosphorylating p300 predicts that the active form of MEKK1 resides in the nucleus, and these two proteins may physically interact. To address this question, we analyzed the localization of endogenous p300 in the presence and absence of either active or inactivate MEKK1. As expected, we find MEKK1⌬ present almost exclusively in the nucleus in a punctate pattern similar to that of the endogenous p300 (compare Fig. 8, E with F). In fact, upon close inspection of the individual and merge images, the active MEKK1 signal appears to overlap with that of p300, suggesting that these two proteins my colocalize in the nucleus. This finding is consistent with a direct regulatory role of MEKK1⌬ on p300 in the nucleus.
Inactive MEKK1 resides predominantly in the cytoplasm (see Ref. 29 and Fig. 8A). As shown in Fig. 8A, HA-tagged full-length MEKK1 (inactive form) is present in both diffused and punctate patterns in the cytoplasm (Fig. 8A). On the other hand, endogenous p300 is present in the nucleus (see Ref. 12, and Fig. 8B, open arrow, indicates an untransfected cell in which p300 is present in the nucleus). Interestingly and strikingly, in the cells that overexpress MEKK1 (Fig. 8, arrowheads), p300 staining shows a significant alteration. In addition to the nuclear staining, it is also present in a punctate pattern in the cytoplasm, identical to the punctate HA-MEKK1 pattern (Fig. 8B, HA-MEKK1transfected cell indicated by a closed arrow). This suggests that some of the p300 molecules may have been retained by HA-MEKK1 in the cytoplasm, perhaps through protein-protein interactions. Taken together, analyses of the subcellular localization of p300 as well as MEKK1 (both active and inactive forms) lend further support to the idea that MEKK1 can regulate p300 activity directly, independent of its downstream JNK kinase.
MEKK1-induced Apoptosis Is Inhibited in a p300-active Ri-bozyme Cell Line-It has been documented that MEKK1 activation can induce cells to undergo apoptosis (29,45,46). To determine whether p300 is involved in the MEKK1 apoptotic pathway, we made use of an MCF-7 cell line that expresses a much reduced level of p300 due to the presence of an active p300 ribozyme (40). As shown previously, these cells, but not MCF-7 cells containing an inactive p300 ribozyme, are deficient in the apoptotic response of cells to ionizing radiation (40). We asked if activated MEKK1 induces apoptosis in MCF-7 cells in a manner that may be dependent on p300. As shown in Fig.  9, compared with the inactive p300 ribozyme cell line which contains wild-type level of p300, the active p300 ribozyme cell line showed a 50% reduction in apoptosis induced by MEKK1⌬ (36% apoptotic cells for inactive ribozyme MCF-7 versus 16% apoptotic cells for p300 active ribozyme MCF-7). This suggests that p300 may be part of the pathway by which MEKK1 induces apoptosis and suggests that the modification of the transcriptional activity of p300 by MEKK1 may be important for this process. sion. In this report, we demonstrate that a kinase in the mitogen-activated kinase pathway, MEKK1, can robustly enhance transcription mediated by p300 when fused to the GAL4 DNA binding domain. This is likely to be physiologically relevant as nocodazole, an extracellular stimulus of MEKK1, can also enhance p300-dependent transcription. MEKK1 has been shown to be important for the apoptotic response of cells to ionizing radiation (29). Significantly, the dynamic interaction between MEKK1 and p300 may be important for MEKK1-induced apoptosis.
MEKK1 has been shown to activate the JNK family of the mitogen-activated kinase members by activating MKK4 (30,31,44), an immediate upstream activator of JNK. Surprisingly, our results show that JNK is not involved in the activation of p300-dependent transcription. First, dominant negative JNK1 does not block MEKK1-induced activation of GAL4-p300-mediated transcription. Second, mutations of all potential JNK sites (proline-directed serines or threonines) within the N2 or C3 region do not impair the ability of MEKK1 to enhance p300mediated transcription. Since MEKK1 can phosphorylate p300 in vitro, these results taken together raise the possibility that MEKK1 may induce p300-mediated transcription by directly phosphorylating p300, particularly within the p300 N2 region. However, since p300 N1 and p300 C3 regions are only weakly phosphorylated by MEKK1 itself (Fig. 7A), it is possible that the ability of MEKK1 to activate GAL4-p300-mediated transcription may also involve additional unidentified MEKK1 downstream kinases. Taken together, our data suggest that stimulation of p300-mediated transcription by MEKK1 may occur via both direct and indirect mechanisms.
Since p300 is a large platform protein and can interact with many different transcription factors, our results cannot rule out the possibility that the effect of MEKK1 on p300 may be mediated by modifications of the interacting transcription factors. For instance, p300 interacts with NFB (47) and MEKK1 activates IB␣ kinase which activates NFB. We thus considered whether MEKK1 merely activated this pathway that facilitated the interaction of NFB with GAL4-p300 and somehow resulted in transcriptional activation. We tested this possibility using the dominant negative form of the IB␣ kinase (34) and found no evidence for the involvement of NFB in the stimulation of p300-mediated transcription by MEKK1 (data not shown).
The ability of MEKK1 to regulate transcription directly is not unprecedented. MEKK1 has been shown to increase the transcriptional activity of GAL4-c-myc and GAL4-Elk-1 (46). The activation of GAL4-c-myc-dependent transcription is believed to also occur in a JNK-independent manner (46). Together with the results described here, these findings suggest that direct phosphorylation of transcriptional regulators by MEKK1 may be a common mechanism by which MEKK1 regulates transcription. Consistent with this idea, we find the active form of MEKK1 (MEKK1⌬) in the nucleus of transfected cells and its localization appears to coincide with that of the endogenous p300 (Fig. 8, E and F). Furthermore, endogenous p300 can be found colocalizing with the punctate cytoplasmic MEKK1 signals in cells overexpressing the full-length inactive form of MEKK1, which resides in the cytoplasm, suggesting a possible physical interaction. Collectively, these findings are consistent with the possibility of direct regulation of p300 by MEKK1.
At present, the phosphoacceptor sites on p300 targeted by MEKK1 or its unknown downstream kinases are not defined. We have mutated all potential JNK sites that consist of proline-directed serines and threonines within p300 aa 302-667 and aa 1709 -1913 without causing alterations in the response to MEKK1. In addition, mutations of serines in a putative MEKK1 phosphorylation motif (DSXXXS) (48) located within p300 aa 302-667 did not abolish the ability of this region to respond to MEKK1 (data not shown). Further studies are necessary to identify the amino acid residues that are either directly or indirectly phosphorylated by MEKK1.
How does phosphorylation of p300 leads to an enhancement of its transcriptional activity? p300 has been shown to interact with members of the basal transcriptional machinery such as TFIIB and TATA-binding protein (4,21,22,26) as well as the RNA polymerase II complex (23). Perhaps one mechanism by which MEKK1 increases the transactivation potential of p300 is to promote the interaction of p300 with these members of the transcriptional basal machinery. An alternative but not mutually exclusive possibility is that phosphorylation of p300 by MEKK1 may enhance the HAT activity of p300. It should be interesting in the future to test directly these possibilities. Activated MEKK1 has been shown to induce cell death (28,45,46,49). We provide evidence here that p300 may be important in mediating MEKK1 induction of the apoptotic pathway in MCF-7 cells. Impaired expression of p300 in MCF-7 cells resulted in significantly lower levels of apoptosis induced by MEKK1 (Fig. 9). As discussed above, the regulation of p300 activity by MEKK1 does not require its downstream kinase JNK. Interestingly, the MCF-7 cell death response to MEKK1 also appears to be JNK-independent (29,46). Collectively, these findings suggest that the ability of MEKK1 to induce apoptosis in MCF-7 cells may involve an increase in the transcriptional activity of p300.
MEKK1-induced apoptosis appears to involve stabilization of p53 as well as an increase in its transcriptional activity (50). Given the fact that p300 is a coactivator for p53 (41,51,52), it is possible that MEKK1-induced phosphorylation of p300 may be important for p300 to serve as a coactivator for p53. It is interesting to note that p300 appears to be essential for the FIG. 9. Effect of p300 ribozyme on the ability of MEKK1⌬ to induce apoptosis in MCF-7 cells. 10 g of pCDNA3 vector, pCDNA3 MEKK1⌬, or pCDNA3 MEKK1⌬K432M were cotransfected along with GFP into MCF-7 cells containing an inactive or active p300 ribozyme. Forty eight hours after transfection, 400 cells were counted, and the percentage of apoptotic cells were scored as described under "Experimental Procedures." Results are representative of two experiments, and the mean Ϯ S.D. of triplicate cultures is indicated. stabilization of p53 as well as increase in p53-dependent transcription in response to ionizing radiation (40). Therefore, it is also possible that phosphorylation of p300 by MEKK1 may contribute to the interaction of p300 with p53 and its stabilization, resulting in changes of p53-dependent gene expression and apoptosis in MCF-7 cells.
In summary, our data show that MEKK1 can significantly enhance transcription mediated by p300. We have identified several regions within p300 that are responsive to MEKK1. The MEKK1-induced transcription mediated by p300 does not appear to require its known downstream kinase JNK. Instead, regulation of p300 activity by MEKK1 may occur via a direct phosphorylation of p300 by MEKK1 and by other currently unknown MEKK1 downstream kinases. Biologically, impaired expression of p300 in MCF-7 cells significantly inhibits the amount of apoptosis induced by MEKK1⌬, suggesting that p300 may be an important component by which MEKK1⌬ induces apoptosis. Our findings thus identify MEKK1 as a kinase that may regulate the transcriptional as well as the biological activities of p300.