Bcr-Abl Regulates Protein Kinase Cι (PKCι) Transcription via an Elk1 Site in the PKCι Promoter*

The protein kinase C (PKC) family of serine/threonine kinases plays an important role in numerous cancer signaling pathways, including those downstream of the bcr-abl oncogene. We demonstrated previously that atypical PKCι is required for Bcr-Abl-mediated resistance of human K562 chronic myelogenous leukemia (CML) cells to Taxol-induced apoptosis. Here, we report that the pattern of PKC isozyme expression characteristic of CML cells is regulated by Bcr-Abl. When Bcr-Abl was expressed in Bcr-Abl-negative HL-60 promyelocytic leukemia cells, expression of the PKCβI, PKCβII, and PKCι genes was induced, whereas expression of the PKCδ gene was reduced to levels similar to those found in CML cells. Given the importance of PKCι in Bcr-Abl-mediated transformation, we characterized the mechanism by which Bcr-Abl regulates PKCι expression. A 1200-bp PKCι promoter construct isolated from genomic DNA was highly active in Bcr-Abl-positive K562 cells and was activated when Bcr-Abl-negative cells were transfected with Bcr-Abl. Bcr-Abl-mediated induction of the PKCι promoter was dependent upon MEK1/2 activity, but not phosphatidylinositol 3-kinase or p38 MAPK activity. Mutational analysis of the PKCι promoter revealed a region between 97 and 114 bp upstream of the transcriptional start site that is responsible for Bcr-Abl-mediated regulation. Mutation of a consensus Elk1-binding site within this region abolished Bcr-Abl-mediated regulation. We conclude that Bcr-Abl regulates PKCι expression through the MEK-dependent activation of an Elk1 element within the proximal PKCι promoter. Our results indicate that Bcr-Abl-mediated transformation involves transcriptional activation of the PKCι gene, which in turn is required for Bcr-Abl-mediated chemoresistance.

Members of the protein kinase C (PKC) 1 family of serine/ threonine kinases participate in many cellular signaling processes, including those involved in cancer development. Specific changes in PKC isozyme expression accompany carcinogenesis in many cancers, including those of the brain, lung, prostate, stomach, skin, breast, liver, small intestine, and colon (reviewed in Ref. 1). Accumulating evidence indicates that changes in PKC isozyme expression contribute to the process of cellular transformation by inducing changes in cellular proliferation, differentiation, motility, and survival. Despite the importance of PKC isozyme regulation in cellular transformation, the transcriptional or translational regulation of PKC isozyme expression has not been well studied in the context of carcinogenesis. Although promoters for several PKC isozymes have been cloned from human and mouse sources (2)(3)(4)(5)(6), only the human PKC␣ promoter has been specifically examined in cancer (2).
Chronic myelogenous leukemia (CML) results from translocation of the abl tyrosine kinase gene from the long arm of chromosome 9 to the breakpoint cluster region of chromosome 22, the Philadelphia chromosome (7). This translocation results in the production of a constitutively active tyrosine kinase fusion gene product, bcr-abl (8). The Philadelphia chromosome is present in 95% of patients with CML, a myeloproliferative disorder that results from the transformation and expansion of a single multipotent hematopoietic stem cell (9). bcr-abl has been shown to be the sole transforming oncogene in animal models of CML (10,11) and is generally considered to be the causative agent in CML carcinogenesis. Multiple breakpoints in the bcr gene result in several forms of bcr-abl. The most common form of bcr-abl in CML is the 210-kDa p210 bcr-abl gene (9). A less common form results when there is a truncated bcr region: 185-kDa p185 bcr-abl . p185 bcr-abl is also transforming and is present in 25% of patients with acute lymphoid leukemia (9,12). Tyrosine kinase activity is necessary for transformation by Bcr-Abl, and a hallmark of Bcr-Abl transformation is extreme resistance to chemotherapeutic agent-induced apoptosis (8,13,14), a phenotype that makes CML difficult to manage clinically.
Bcr-Abl has pleiotropic effects on multiple signal transduction pathways (reviewed in Ref. 15). Bcr-Abl is known to activate Ras/MEK (16,17), phosphatidylinositol (PI) 3-kinase-dependent pathways (18,19), and JAK (Janus kinase)/STAT (signal transducer and activator of transcription) (20,21), and the hyperproliferative phenotype conferred by Bcr-Abl appears to require contributions from each of these pathways (15). The extreme resistance to apoptosis associated with Bcr-Abl-mediated transformation is thought to be specifically dependent on the Ras/MEK and PI 3-kinase pathways (22)(23)(24)(25)(26). Our group has demonstrated previously that several PKC isozymes are important in the biology of human K562 CML cells (27)(28)(29)(30). We have shown that PKC␤II is required for CML cell proliferation, whereas PKC is necessary for Bcr-Abl-mediated resistance to antineoplastic drug-induced apoptosis (27,29,30). We have also shown that PKC is a critical regulator of the NF-B pathway and that NF-B activity is necessary for the antiapoptotic effect of Bcr-Abl (28). PKC␤II appears to function downstream of Bcr-Abl and is necessary for Bcr-Abl-dependent phosphorylation of the proto-oncogene FUS (heterogeneous nuclear ribonucleoprotein-1) (31). This phosphorylation prevents proteolysis of the FUS protein, which is thought to mediate Bcr-Abl-dependent proliferation and lack of differentiation (31).
Because of the strong link between specific PKC isozymes and Bcr-Abl-mediated transformation, we profiled the expression of PKC isozymes as a consequence of expression of Bcr-Abl using Affymetrix gene arrays. We found that Bcr-Abl regulates expression of multiple PKC isozymes implicated in Bcr-Ablmediated transformation, including PKC␤II and PKC. We demonstrate that Bcr-Abl regulates PKC expression through MEK-dependent activation of a consensus Elk1 element within the proximal promoter of the PKC gene.
High Density Oligonucleotide Array Analysis-RNA samples from HL-60/Neo and HL-60/bcr-abl cells were hybridized to Hu95A Gene-Chips (Affymetrix, Inc., Santa Clara, CA) containing 12,626 sequenced genes essentially as we described previously (32). GeneChip arrays were scanned using a gene array scanner (Hewlett-Packard Co.) and analyzed using GeneChip Analysis Suite 4 software (Affymetrix, Inc.).
Real-time Reverse Transcriptase-Polymerase Chain Reaction Analysis of Gene Expression-Real-time reverse transcriptase-PCR was conducted using TaqMan technology on an Applied Biosystems 7000 sequence detection system. A 20ϫ assay mixture of primers and TaqMan minor grove binder (MGB) probes (6-phosphoramidite 6-FAM™ dyelabeled) was used for each target gene and the endogenous control, human glyceraldehyde-3-phosphate dehydrogenase. Primers were designed to span exon-exon junctions so as not to detect genomic DNA. The primer and probe sequences were as follows. Human PKC probe, CCTTCCTTACACATGCC; forward primer, GAAGGCCACATTAAACT-CACTGACTAC; and reverse primer, GGTTGTATCTCCTGGCCGT-AAT. Human PKC␤I probe, TCTTGTCTCTAGCTTTTG; forward primer, AAACTTGAACGCAAAGAGATCCA; and reverse primer, CTG-TCTGGTGAACTCTTTGTCGAA. Human PKC␤II probe, TCGCCCAC-AAGCT; forward primer, AAACTTGAACGCAAAGAGATCCA; and reverse primer, ATCGGTCGAAGTTTTCAGCATT. Human PKC␦ probe, CCCACTGTTGTCTTGC; forward primer, TTCGAGAAGAAGACCGG-AGTTG; and reverse primer, CTCCCAGATCTTGCCGTAGGT. Primer sequences were searched against the Celera Database to confirm specificity. The method was performed following the manufacturer's instructions (Applied Biosystems) and has been described previously (32).
Plasmids and Constructs-The p210 bcr-abl , p185 bcr-abl , and p185 kd-bcr-abl (where kd is kinase-dead) expression plasmids were constructed by excision of the Bcr-Abl coding regions from pEXV retroviral expression vectors using non-directional cloning via an EcoRI site (a gift from Dr. A. M. Pendergast, Duke University) and ligation into either pCMV-Tag2b (p210 bcr-abl ; Stratagene) or pCMV-Script (p185 bcr-abl and p185 kd-bcr-abl ) for transient expression in mammalian cells. Human genomic P1-derived artificial chromosome (PAC) clones for PKC promoter cloning were obtained from Genome Systems Inc. using the 5Ј-most 500 bp of the PKC coding sequence as a probe for PAC clone selection. All genomic PKC promoter constructs were amplified by PCR directly from the genomic PAC clone and cloned by TOPO-TA cloning (Invitrogen) into the pCR2.5 vector. Each 5Ј-primer contained a 5Ј-XhoI site, and each 3Ј-PCR primer contained a 3Ј-HindIII restriction site to facilitate subsequent cloning. A two-step PCR procedure was used to generate site-directed mutants. Briefly, 5Ј-and 3Ј-overlapping regions were amplified from the original PAC clone, melted, mixed, and then re-amplified to produce a full-length, mutated, finished product, which was subsequently cloned into TOPO-TA as described above. The promoter constructs were directionally cloned into the pGL3 reporter plasmid (Promega). The Renilla luciferase reporter plasmid pRLSV40 (Promega) was used as an internal control. All plasmids were purified by ion-exchange chromatography (QIAGEN Inc.), resolved by agarose electrophoresis, and confirmed by sequencing. All constructs used the reverse primer sequence 5Ј-CGGCTCCTCCCACTGCCCGT-3Ј. The forward primer sequences were as follows: Ϫ1200, TCGCCCAGGCTG-GAGTGCAG; Ϫ166, GTTCAGCCCGGTATTGAGGC; Ϫ141, GAGCTA-TGACTTAGGTTGCG; Ϫ97, CTTTGGAGCGAGCGAAGTGG; 141Mut1, GAGCTATGACTTAGGTTGCGCCCGCCCTA; 141Mut2, GAGCTGGC-TGGATGGTTGCGATGCTTATACCCCGGAATGTGGG; 141Mut3, GA-GCTGGCTGGATGGTTGCGCCCGCCCTAACATCTAATGTGGGCCT-TTG; 166Mut1, ACCGTCCCCGGTATTGAGGCTCCTTG; and 166-Mut2, GTTCAGCCCGGTATTGTCATGACTTGAGCTGGCTGGATGG-TTG. For two-step PCR, the external forward primer was the Ϫ166 primer above. The internal primers were as follows: 166Mut3, GCCC-ACATGAGACATCAGGGCGGGCGCAACCATCCA (internal reverse) and CCGCCCTGATGTCTCATGTGGGCCTTTGGAGCGAGCGAA-GTGG (internal forward); 166Mut4, AAAGGCCATGTCGAGACGGTA-GGGCGGGCGCAACCATCC (internal reverse) and CTACCGTCTCGA-CATGGCCTTTGGAGCGAGCGAAGTGG (internal forward); and 166⌬-Elk, CTTCGCTCGCTCCAAAGGCCGGGCGGGCGCAACCA (internal reverse) and TGGTTGCGCCCGCCCGGCCTTTGGAGCGAG (internal forward).
Transfections and Luciferase Reporter Assays-Transient transfections were carried out using the LipofectAMINE Plus transfection reagent (Invitrogen) as described by the manufacturer with minor modifications. Suspension cultures of K562 cells in log-phase growth were plated in 200 l of serum-free medium (5 ϫ 10 5 cells)/well of a 24-well plate for transfection. Adherent cultures (HEK293 and HepG2 cells) were plated to 70% confluence 24 h prior to transfection, and the growth medium was replaced with serum-free medium immediately before transfection. Transfections were carried out in triplicate using pGL3 reporter plasmid DNA containing PKC promoter constructs at 0.2 g/well. In some cases, Bcr-Abl reporter or empty vector DNA at 0.2 g/well was included in the transfection mixture. All transfections contained the control Renilla pRLSV40 plasmid at 25 ng/well as an internal control. Cellular protein was isolated in 100 l of passive lysis buffer (Promega) 24 or 48 h after transfection as indicated in the figure legends. Firefly and Renilla luciferase activities were measured in 10 l of cell lysate using the dual luciferase assay kit (Promega) on a Monolight 2010 luminometer.

Expression of Bcr-Abl in HL-60 Cells Confers
Chemoresistance to Taxol-HL-60 cells are promyelocytic leukemia cells that lack the 9:22 chromosomal translocation that results in expression of the Bcr-Abl fusion protein. We demonstrated previously that parental HL-60 cells are sensitive to Taxolinduced apoptosis (27). In contrast, K562 cells express Bcr-Abl and are highly resistant to Taxol-induced apoptosis (27). We utilized HL-60 cells that have been stably transfected with either a control neomycin resistance plasmid (HL-60/Neo cells) or the bcr-abl oncogene (HL-60/bcr-abl cells) (33) to assess whether Bcr-Abl induces a Taxol-resistant phenotype in HL-60 cells. HL-60/bcr-abl, HL-60/Neo, and K562 cells were treated with 100 nM Taxol for 24 h, and apoptosis was measured by the presence of apoptotic bodies in 4,6-diamidino-2-phenylindolestained cells (Fig. 1A) as described previously (27). As expected, HL-60/Neo cells were highly sensitive to Taxol and rapidly underwent apoptosis in response to Taxol. In contrast, HL-60/ bcr-abl cells were highly resistant to Taxol-induced apoptosis and, much like K562 cells, underwent mitotic arrest (but not apoptosis) when treated with Taxol. Therefore, Bcr-Abl confers resistance to Taxol-induced apoptosis when expressed in HL-60 cells, consistent with the original characterization of HL-60/ bcr-abl cells (33).
Bcr-Abl Induces Changes in PKC Isozyme Gene Expression-We demonstrated previously the importance of multiple PKC isozymes in Bcr-Abl-mediated transformation, including resistance to Taxol-induced apoptosis (27,30). Therefore, we next examined the effect of Bcr-Abl expression on PKC isozyme expression in HL-60/bcr-abl cells by GeneChip microarray analysis (Fig. 1B). Our analysis revealed a significant increase in expression of the PKC␤I, PKC␤II, and PKC genes and a significant decrease in expression of the PKC␦ gene in HL-60/ bcr-abl cells compared with HL-60/Neo cells. In contrast, ex-pression of the PKC⑀, PKC, PKC␥, PKC, PKC, and PKC genes was not significantly changed in HL-60/bcr-abl cells compared with HL-60/Neo cells.
We confirmed the changes in expression of PKC␤I, PKC␤II, PKC␦, and PKC by real-time reverse transcriptase-PCR (Fig.  2). Expression of the mRNA for each PKC isozyme was measured in K562 cells ( Fig. 2A), HL-60/Neo cells (Fig. 2B), and HL-60/bcr-abl cells (Fig. 2C) and, in each case, normalized to PKC mRNA levels. K562 and HL-60/bcr-abl cells exhibited a strikingly similar pattern of PKC expression characterized by relatively high PKC, PKC␤I, and PKC␤II levels and low PKC␦ levels. In contrast, Bcr-Abl-negative HL-60/Neo cells exhibited a distinct pattern characterized by relatively low PKC, PKC␤I, and PKC␤II levels and high PKC␦ levels. To compare PKC isozyme levels in all three cell lines, the level of each isozyme is expressed relative to the level of that isozyme in K562 cells (Fig. 2D). This analysis demonstrated a dramatic increase in the levels of PKC␤I, PKC␤II, and PKC mRNAs and a decrease in the levels of PKC␦ mRNA in both K562 and HL-60/bcr-abl cells compared with HL-60/Neo cells. These data are consistent with our microarray data and confirm the changes in expression of PKC isozymes by Bcr-Abl. Furthermore, they demonstrate that expression of Bcr-Abl induces a pattern of PKC isozyme expression strikingly similar to that in K562 CML cells.
We next confirmed that the changes in PKC isozyme mRNA expression induced by Bcr-Abl result in concordant changes in PKC isozyme protein levels. For this purpose, we assessed the levels of PKC␤I, PKC␤II, PKC␦, and PKC in HL-60/Neo, HL-60/bcr-abl, and K562 cells by immunoblot analysis (Fig. 3). In each case, the level of PKC isozyme protein expression was consistent with the observed change in the mRNA level for that isozyme as demonstrated by microarray and real-time reverse transcriptase-PCR analysis, demonstrating the similarities in PKC isozyme expression in K562 CML cells and HL-60 cells engineered to express Bcr-Abl.
Bcr-Abl Regulates PKC Gene Promoter Activity-We demonstrated previously a requisite role for PKC in Bcr-Abl-mediated resistance to apoptosis (27,30). Therefore, we focused our attention on the regulation of PKC transcription by Bcr-Abl. We began our analysis by cloning and characterizing the upstream regulatory promoter region of the PKC gene. Two human PAC clones were isolated based on hybridization to a probe consisting of the 5Ј-most end of the PKC mRNA sequence (Genome Systems Inc.). Comparison of the PKC mRNA 5Ј-probe used for the PAC clone isolation with the NCBI published draft of the human genome using the BLAST algorithm revealed a large (Ͼ14 kb) region of known genomic sequence upstream of the PKC coding region. PCR primers complementary to the genomic PKC DNA sequence were used to clone two PAC fragments encompassing 3929 and 1200 bp immediately upstream of the transcriptional start site of the PKC gene. These two clones were inserted into the pGL3 reporter plasmid upstream of a luciferase reporter gene and assessed for promoter activity. The 1200-bp promoter construct contains promoter activity that is approximately three times greater compared with a control SV40 promoter construct when transfected into K562 cells (Fig. 4A). Furthermore, both the 3929-and 1200-bp PKC promoter constructs were highly activated by Bcr-Abl (ϳ5-fold) when cotransfected with a mammalian expression plasmid containing FLAG-tagged p210 bcr-abl for 48 h into the Bcr-Abl-negative cell line, HepG2 (Fig. 4B). Similar results were obtained when Bcr-Abl and the PKC promoter constructs were cotransfected into HEK293 cells, indicating that Bcr-Abl-mediated activation of the PKC promoter is not specific to HepG2 cells (data not shown). Thus, the 1200-bp PKC promoter construct exhibits promoter activity and contains elements responsible for p210 bcr-abl -mediated promoter activation.
We next assessed whether the 1200-bp PKC promoter construct is also activated by the p185 variant of the bcr-abl oncogene and whether activation of the promoter requires tyrosine kinase activity of Bcr-Abl. HEK293 cells were cotransfected with the 1200-bp PKC promoter construct and a mammalian expression plasmid (pCMV) containing p210 bcr-abl , p185 bcr-abl , or a kinase-deficient p185 bcr-abl mutant (p185 kd-bcr-abl ) (Fig. 4C). A similar level of promoter activation was observed in the presence of either p210 bcr-abl or p185 bcr-abl , whereas p185 kd-bcr-abl was unable to activate the 1200-bp PKC promoter construct. Immunoblot analysis using an antibody against the Abl region common to p210 bcr-abl , p185 bcr-abl , and p185 kd-bcr-abl demonstrated that similar levels of the three Bcr-Abl proteins were expressed in the cells used in these experiments (data not shown). These data demonstrate that both p210 bcr-abl and p185 bcr-abl can activate the PKC promoter and that promoter activation is dependent upon Bcr-Abl tyrosine kinase activity.
Bcr-Abl-mediated PKC Promoter Activity Is MEK-dependent-Bcr-Abl mediates its cellular effects through activation of several major signaling pathways. We employed selective chemical inhibitors of specific signaling components to determine which pathway(s) is involved in Bcr-Abl-mediated activation of the PKC promoter (Fig. 5). Cotransfection of p210 bcr-abl and the 1200-bp PKC promoter construct into HepG2 cells led to a 5-6-fold activation of the PKC promoter compared with cells expressing the PKC promoter and an empty vector control. Bcr-Abl-mediated promoter activation was not altered by the presence of the PI 3-kinase inhibitors wortmannin and LY294002 (Fig. 5A). Likewise, selective inhibitors of p38 MAPK (SB202190 and SB203580) had no effect on Bcr-Abl-mediated activation of

FIG. 3. Bcr-Abl induces changes in PKC isozyme protein expression.
Total cell lysates from K562, HL-60/Neo, and HL-60/bcr-abl cells in the logarithmic growth phase were subjected to immunoblot analysis for Bcr-Abl, PKC␤I, PKC␤II, PKC␦, PKC, and actin using antibodies specific to antigen as described under "Experimental Procedures." the PKC promoter (Fig. 5B). However, Bcr-Abl-mediated activation of the PKC promoter was significantly inhibited by the MEK1/2-selective inhibitor U0126, but not by the inactive analog U0124 (Fig. 5C). Therefore, Bcr-Abl-mediated activation of the PKC promoter is dependent upon MEK1/2 activity, but is independent of p38 and PI 3-kinase activity.

FIG. 4. Identification of a functional PKC promoter that is activated by Bcr-Abl.
A, 1200 bp of sequence 5Ј to the PKC coding sequence were isolated and cloned into the pGL3-luciferase reporter vector to produce the 1200-bp PKC promoter construct. This construct was transfected into K562 cells and assessed for promoter activity by measuring luciferase activity 24 h after transfection as described under "Experimental Procedures." PKC promoter activity is normalized to the activity of a Renilla SV40 plasmid that was cotransfected as an internal control for transfection efficiency. Error bars represent the S.D. of three replicate determinations. *, p Ͻ 0.02. B, a PKC promoter construct corresponding to either 1200 or 3929 bp of sequence 5Ј to the coding region of PKC or a pGL3 empty plasmid was transfected into HepG2 cells along with either a pCMV-Tag2b empty vector or the Tag2b/p210 bcr-abl vector. Cells were harvested after 48 h of transfection and analyzed for PKC promoter activity as described under "Experimental Procedures." Results are expressed as the -fold induction of the indicated PKC promoter construct in the presence of Bcr-Abl. Values are normalized to Renilla SV40 as an internal control for transfection efficiency. Error bars represent the S.D. of three replicate determinations. C, the 1200-bp PKC promoter construct was cotransfected with Tag2b, Tag2b/p210 bcr-abl , pCMV, pCMV-p185 bcr-abl , or pCMV-p185 kd-bcr-abl into HepG2 cells. Cells were harvested 48 h after transfection and assessed for PKC promoter activity as described under "Experimental Procedures." Results are expressed as -fold induction in the presence of Bcr-Abl. Values are normalized to Renilla SV40 as an internal control for transfection efficiency. Error bars represent the S.D. of three replicate determinations. *, p Ͻ 0.002 between p185 bcr-abl and p185 kd-bcr-abl ; **, p Ͻ 0.002 between p185 and p185 kinase dead.

FIG. 5. Activation of the PKC promoter by Bcr-Abl is dependent upon MEK1/2, but not PI 3-kinase or p38 MAPK.
The 1200-bp PKC promoter construct was cotransfected with either Tag2b or p210 bcr-abl into HepG2 cells, and PKC promoter activity was assessed in the presence or absence of chemical inhibitors as described under "Experimental Procedures." A, transfectants were treated with the PI 3-kinase inhibitors wortmannin (500 nM) and LY294002 (10 M) or with an equal volume of dimethyl sulfoxide (DMSO). B, transfectants were treated with the p38 MAPK inhibitors SB202190 and SB203580 or the negative control SB202474 at 20 M or with diluent dimethyl sulfoxide.

Identification of a Consensus Elk1 Element in the PKC Promoter That Confers Bcr-Abl
Responsiveness-We next determined the region(s) of the PKC promoter responsible for Bcr-Abl-mediated activation. 5Ј-Deletions of the existing 3929-bp promoter sequence corresponding to 1200, 166, and 97 bp of the proximal PKC promoter were generated by PCR, cloned into the pGL3-luciferase reporter plasmid, transfected into Bcr-Ablpositive K562 cells, and assessed for Bcr-Abl-mediated activation (Fig. 6A). Promoter activity remained high until region Ϫ166 to Ϫ97 was deleted, whereupon promoter activity was dramatically reduced. Thus, important transcriptional regula-tory elements are present in region Ϫ166 to Ϫ97 of the PKC promoter.
Several consensus transcription factor-binding motifs were detected within region Ϫ166 to Ϫ97 of the PKC promoter using the TRANSFAC promoter analysis algorithm (Fig. 6B). The presence of an Elk1 element within this region was of particular interest because Elk1 is known to be a critical downstream component of the Ras/MEK1/ERK transcription pathway activated by Bcr-Abl (34), a pathway that is also important in Bcr-Abl-mediated transformation. Furthermore, our inhibitor studies demonstrated that Bcr-Abl-mediated activation of the FIG. 6. Identification of a region within the PKC promoter that confers Bcr-Abl responsiveness. A, PKC promoter deletion constructs were produced as described under "Experimental Procedures." K562 cells were transfected with equal amounts of each PKC promoter construct, and cell lysates were harvested after 24 h for analysis of PKC promoter activity. Values are normalized to Renilla SV40 as an internal control for transfection efficiency. Error bars represent the S.D. of three replicate determinations. *, p Ͻ 0.002 between the 166-and 97-bp promoter constructs. B, shown is a schematic diagram of the proximal PKC promoter between Ϫ166 and Ϫ1 upstream of the coding region indicating potential transcription factor response elements. HNF-4, hepatocyte nuclear factor-4; MRF-2, myogenic regulatory factor-2; LEF-1, lymphoid enhancer factor-1; TF, transcription factor. PKC promoter requires MEK1/2 activity, suggesting that an Elk1 element may be involved in Bcr-Abl-mediated activation of the PKC promoter.
The possible involvement of the Ϫ114 consensus Elk1-binding site in Bcr-Abl-induced PKC promoter activity was assessed by generating a series of randomized and deletion mutations within region Ϫ166 to Ϫ97 of the PKC promoter (Fig.  7A). Mutations that disrupt the Elk1 site (166Mut3, 166Mut4, 166⌬Elk, 141Mut2, and 141Mut3) invariably resulted in a significant reduction in promoter activity, indicating that the Elk1-binding element within the PKC promoter is necessary for transcriptional activation in Bcr-Abl-positive K562 cells.
To determine whether the Elk1 element is activated by Bcr-Abl, the 166-bp, 166Mut4, and 97-bp PKC promoter constructs or a control pGL3 reporter construct was cotransfected with either p210 bcr-abl or an empty vector into HepG2 cells (Fig. 7B). The 166-bp PKC promoter construct was activated by Bcr-Abl to a level comparable with that observed with the original 1200-bp PKC promoter construct. However, the 166Mut4 PKC promoter construct exhibited reduced responsiveness to Bcr-Abl to a level indistinguishable from that observed with the 97-bp promoter construct, indicating that the Elk1 element is necessary for full induction of the PKC promoter by Bcr-Abl. DISCUSSION The PKC family of protein kinases has been implicated in a variety of cellular functions, including the control of cellular proliferation, differentiation, and survival. Since the discovery that PKC is the major receptor for the tumor-promoting phorbol esters, PKC has been implicated in cancer. Reproducible changes in PKC isozyme expression have been demonstrated in cancers from many different tissue sites, suggesting that specific PKC isozyme expression patterns may be involved in cancer development and progression (1). Accumulating evidence suggests that at least some of these changes in PKC isozyme expression are critical to tumor promotion and progression (1). However, despite the emerging recognition that PKC isozymes are critical players in cancer, relatively little is known about how PKC isozyme expression in regulated in either the physiologic or pathologic state.
Our laboratory has demonstrated that two PKC isozymes (PKC␤II and PKC) are critical in some forms of cancer, including colon cancer (35)(36)(37) and CML (27,29,30). Our early studies demonstrated that PKC␤II is required for K562 cells to proliferate in culture (29). We further demonstrated that PKC␤II mediates mitotic phosphorylation of the major nuclear envelope protein lamin B at sites required for nuclear lamina disassembly at the time of mitosis, providing a plausible mechanism by which PKC␤II promotes cellular proliferation of CML cells (38 -40). In addition, PKC␤II is important for Bcr-Ablmediated phosphorylation and stabilization of the FUS protooncogene (31), an activity required for Bcr-Abl-mediated proliferation. Interestingly, both FUS and the lamins are nuclear proteins and, at least in the case of the lamins, are direct substrates for PKC␤II-mediated phosphorylation (38 -40). These results are consistent with the notion that nuclear activation of PKC␤II is an important determinant for the proliferative effects of PKC␤II in CML cells.
PKC plays a distinct, but equally important role in CML biology. We demonstrated previously that PKC is required for Bcr-Abl-mediated resistance of K562 cells to Taxol-induced apoptosis (27,30). Acquired resistance to multiple chemotherapeutic agents is a hallmark of CML, making these tumors difficult to manage clinically. This acquired resistance is dependent upon both Bcr-Abl and PKC (27,30). K562 cells can be sensitized to chemotherapy-induced apoptosis by inhibiting PKC expression or activity (27,30). In more recent studies, we demonstrated that Bcr-Abl requires NF-B activation to exert its anti-apoptotic phenotype and that PKC is necessary downstream of Bcr-Abl and upstream of NF-B activation in a key anti-apoptotic pathway in K562 cells (28).
In this study, we have provided the first direct evidence that expression of a single transforming oncogene (bcr-abl) regulates the pattern of PKC isozyme expression in CML cells. Using gene profiling analysis, we demonstrated that expression of Bcr-Abl in Bcr-Abl-negative HL-60 leukemia cells leads to a pattern of PKC isozyme expression strikingly similar to that observed in Bcr-Abl-positive K562 CML cells. The changes induced by Bcr-Abl are consistent with the known function of individual PKC isozymes in CML biology. Thus, Bcr-Abl induces expression of PKC␤II and PKC, which are critical for Bcr-Abl-mediated proliferation and chemoresistance, respectively. Furthermore, we demonstrated that expression of PKC␦ is repressed by Bcr-Abl. This is consistent with the anti-apoptotic phenotype induced by Bcr-Abl because PKC␦ has been shown to be a pro-apoptotic gene (reviewed in Ref. 41). The fact that the pattern of PKC isozyme expression induced by Bcr-Abl in promyelocytic leukemia (HL-60) cells is so similar to that observed in K562 cells suggests that Bcr-Abl is a major determinant for PKC isozyme patterning in CML cells.
Because of the importance of PKC activity in Bcr-Abl-mediated resistance to apoptosis, we characterized the mechanism by which Bcr-Abl regulates PKC expression. Our results demonstrate that the region immediately upstream of the known transcriptional start site through the Ϫ1200-bp region constitutes a viable PKC promoter that is highly inducible by exogenously expressed p210 bcr-abl and p185 bcr-abl . Our data demonstrate that Bcr-Abl-mediated activation of PKC promoter activity requires Bcr-Abl tyrosine kinase activity, consistent with the requirement of kinase activity for Bcr-Abl transformation and chemoresistance. Mutational analysis of the PKC promoter localized the major target for Bcr-Abl-mediated activation to a consensus Elk1 site within the proximal region of the PKC promoter. Disruption of this site within the PKC promoter abolishes Bcr-Abl-mediated activation. Because Bcr-Abl induces both the PKC and PKC␤ genes, it is possible that Bcr-Abl induces PKC␤ transcription through an Elk1-dependent mechanism. Inspection of a 2.3-kb promoter construct cloned from the human PKC␤ gene revealed no consensus Elk1 site, although related Ets-like elements are present. Whereas this does not preclude the possibility of an Elk1-dependent mechanism in Bcr-Abl-mediated regulation of PKC␤ expression, we have no evidence at present to suggest such a common mechanism. Studies are currently underway to determine whether the human PKC␤ promoter is responsive to Bcr-Abl expression and, if so, what promoter elements are responsible.
The Elk1 transcription factor is a known downstream target for Ras/MEK-mediated transcriptional regulation (34), suggesting the involvement of the Ras/MEK/ERK pathway in the activation of PKC promoter activity by Bcr-Abl. Consistent with this finding, we have shown that induction of PKC promoter activity by Bcr-Abl is blocked by the MEK1/2-selective inhibitor U0126. These results are consistent with the observation that MEK1/2 activity is important for Bcr-Abl-mediated resistance to apoptosis (24,42,43) and that Bcr-Abl-mediated activation of NF-B/RelA requires both Bcr-Abl kinase activity and Ras activation (44,45). Our recent studies demonstrated that PKC is downstream of Bcr-Abl and upstream of NF-B in an apoptotic resistance pathway in CML cells (28).
When combined with the known signaling properties of Bcr-Abl, our data led us to propose the following pathway for Bcr-Abl induction of PKC (Fig. 8). By virtue of its constitutively active tyrosine kinase activity, Bcr-Abl activates MEK1/2 kinase through a Ras-dependent pathway. It has been shown that autophosphorylation of tyrosine 177 on Bcr-Abl recruits Grb2 to Bcr-Abl, leading to activation of the canonical Ras/ MEK pathway (46). MEK activates ERK, which subsequently phosphorylates and activates the Elk1 transcription factor. Activated Elk1 then stimulates PKC promoter activity through a consensus Elk1 response element located 114 bp upstream of the PKC transcriptional start site. Transcriptional activation leads to elevated PKC protein expression, which is required for Bcr-Abl-mediated resistance to apoptosis (27,30). PKC acts to protect cells from Taxol-induced apoptosis by activating NF-B by two distinct mechanisms (28). First, PKC stimulates degradation of IB and nuclear translocation of NF-B (28). In addition, PKC induces transactivation of nuclear NF-B (28). NF-B is known to induce expression of many anti-apoptotic proteins, including hsp70, IAP (inhibitor of apoptosis), Bcl-2, and Bax. This proposed pathway for apoptotic resistance in K562 cells has several implications for therapeutic intervention in CML. Currently, the only known cure for CML is bone marrow transplantation; and until fairly recently, there were few effective chemotherapeutic modalities to manage the disease. The Bcr-Abl tyrosine kinase inhibitor imatinib recently met with great success in initial clinical trials, showing dramatic initial responses (47). However, imatinib-resistant CMLs are rapidly emerging as treatment with imatinib becomes more prevalent, and patients removed from imatinib treatment often relapse (47). Drugs specifically targeted to the PKC-dependent apoptotic resistance pathway induced by Bcr-Abl might be of therapeutic benefit in combination with existing therapy with imatinib. Additionally, inhibition of PKC transcription or PKC activity could potentially inhibit Bcr-Abl-mediated apoptotic resistance while having little or no effect on Bcr-Abl-mediated hyperproliferation. Such treatment could generate CML cells that are more sensitive to proliferation-dependent chemotherapeutic drugs such as Taxol. In this regard, specific inhibitors of MEK1/2 have been shown to enhance the effects of other chemotherapeutic drugs (42). Specific inhibition of PKC may be even more effective than inhibition of MEK for combination chemotherapy because PKC appears to be important for Bcr-Abl-mediated apoptotic resistance, but not for cellular proliferation (27,30), whereas MEK appears to play a role in both the chemoresistance and proliferation induced by Bcr-Abl. Our results indicate that PKC is an attractive therapeutic target for Bcr-Abl-positive leukemias.