Activated α2-Macroglobulin Regulates Transcriptional Activation of c-MYC Target Genes through Cell Surface GRP78 Protein*

Activated α2-macroglobulin (α2M*) signals predominantly through cell surface GRP78 (CS-GRP78) to promote proliferation and survival of cancer cells; however, the molecular mechanism remains obscure. c-MYC is an essential transcriptional regulator that controls cell proliferation. We hypothesize that α2M*/CS-GRP78-evoked key signaling events are required for transcriptional activation of c-MYC target genes. Activation of CS-GRP78 by α2M* requires ligation of the GRP78 primary amino acid sequence (Leu98–Leu115). After stimulation with α2M*, CS-GRP78 signaling activates 3-phosphoinositide-dependent protein kinase-1 (PDK1) to induce phosphorylation of PLK1, which in turn induces c-MYC transcription. We demonstrate that PLK1 binds directly to c-MYC and promotes its transcriptional activity by phosphorylating Ser62. Moreover, activated c-MYC is recruited to the E-boxes of target genes FOSL1 and ID2 by phosphorylating histone H3 at Ser10. In addition, targeting the carboxyl-terminal domain of CS-GRP78 with a mAb suppresses transcriptional activation of c-MYC target genes and impairs cell proliferation. This work demonstrates that α2M*/CS-GRP78 acts as an upstream regulator of the PDK1/PLK1 signaling axis to modulate c-MYC transcription and its target genes, suggesting a therapeutic strategy for targeting c-MYC-associated malignant progression.

CS-GRP78 is a multifunctional receptor that forms complexes with phosphatidylinositol 3-kinase (PI3K) and enhances phosphatidylinositol 3,4,5-trisphosphate production, consistent with its novel role as a regulator of the PI3K/Akt signaling pathway. Thus it promotes cell proliferation, survival, metastasis, and chemoresistance (9, 20 -22). CS-GRP78, through its NH 2 -terminal domain, drives PI3K/Akt activity (2), whereas targeting the COOH-terminal domain with antibody promotes apoptotic signaling (21,23). Recently, we demonstrated in vivo that targeting the GRP78 COOH-terminal domain with monoclonal antibody C38 (C38 mAb) delays tumor growth and prolongs survival (15). We also demonstrated that ␣ 2 M*/CS-GRP78 signaling is required for mechanistic target of rapamycin (mTOR) complex-mediated phosphorylation of Akt by 3-phosphoinositide-dependent protein kinase-1 (PDK1) (22). PDK1 regulates a diversity of substrates and targets that induce aberrant signaling in human malignancy (24). Indeed, recent studies show that PDK1 is required for c-MYC accumulation, and it regulates c-MYC activity through the downstream target PLK1 (25), indicating a potential functional link of ␣ 2 M*/CS-GRP78 signaling and c-MYC in proliferation of cancer cells. ␣ 2 M*/CS-GRP78-mediated PI3K/Akt signaling is well documented; however, its role in cancer-associated gene regulation by transcription factors has yet to be identified.
The oncogene c-MYC globally reprograms cells and drives proliferation by regulating an estimated 15% of the genes in the human genome (26). Recent work suggests that rather than acting as a general amplifier of transcription (27,28) c-MYC activates and represses transcription of discrete gene sets, leading to changes in cell proliferation, tumor progression, and maintenance (29). Moreover, phosphorylation of c-MYC at certain sites governs its activation and subsequent biological functions through transcriptional activation of target genes that are necessary for cell proliferation. Specifically, Ser 62 phosphorylation is necessary for its oncogenic activity (30). A key question is whether ␣ 2 M*/CS-GRP78 signaling is required for activation of c-MYC and its downstream target genes.
In the present study, we demonstrate that ␣ 2 M*/CS-GRP78mediated PDK1/PLK1 signaling contributes to the transcriptional activation of c-MYC target genes and proliferation. We further demonstrate that PLK1 can directly bind to c-MYC and promote its transcriptional activity by phosphorylating at his-tone H3 Ser 10 (H3S10). These findings suggest that ␣ 2 M*/CS-GRP78 signaling drives c-MYC target gene expression in human cancers and provide a therapeutic approach for targeting c-MYC-driven tumors.

Experimental Procedures
Cell Culture-1-LN prostate cancer cells were a kind gift from Dr. Philip Walther, Department of Surgery, Duke University Medical Center. They now reside in our laboratory and are available on request. DU145 prostate cancer cells, A375 melanoma cells, and U373 glioma cells were purchased from the Duke Cell Culture Facility. 1-LN and DU145 cells were maintained in RPMI 1640 medium (Sigma) containing 10% fetal bovine serum (FBS), 1% penicillin/streptomycin at 37°C in a 5% CO 2 -humidified atmosphere. A375 and U373 cells were maintained in DMEM (high glucose; Gibco, Life Technologies) containing 10% FBS, 1% penicillin/streptomycin at 37°C in a 5% CO 2 -humidified atmosphere.
Small Interfering RNA (siRNA) Interference and Lentiviral Transfections-siRNAs targeting c-MYC and nonspecific siRNA were transfected into 1-LN cells with Lipofectamine 2000 reagent according to the manufacturer's instructions. shc-MYC lentiviral particles (Clone ID TRCN000000-39642) were obtained from Sigma and transfected into 1-LN cells according to the manufacturer's instructions. After transfection with sh-c-MYC vector, 1-LN cells were selected with 2 mg/ml puromycin.
Flow Cytometry-CS-GRP78 was analyzed by flow cytometry as described previously (15). The mean fluorescence intensity of the signal was calculated by Flow Jo software, and signal obtained from GRP78 was normalized with that obtained from isotype controls.
Cell Proliferation Assays-Cells were plated in 96-well plates at 10,000 cells/well in 0.1 ml of growth medium containing inhibitor BX795, BI2536, or C38 mAb for 72 h. Cell viability was measured according to the manufacturer's instruction protocol by using the XTT assay. Absorbance was read at ϭ 450 nm.
Soft Agar Colony Formation Assay-Soft agar assays were conducted in 6-well plates in triplicate. For each well, 10,000 cells were mixed in growth medium containing 1.2% agarose and inhibitor BX795, BI2536, or C38 mAb. Cells were then layered over 2% agarose in regular medium. Medium containing BX795, BI2536, or C38 mAb was added to each well every 3 days. The assays were terminated after 21 days, colonies were counted under a microscope or stained with crystal violet, and pictures were taken.
Immunoblotting and Immunoprecipitation-Protein extracts, immunoblotting, and immunoprecipitate analysis were performed as described previously (27), and all blots are representative of a minimum of two independent experiments.
Immunoprecipitation and in Vitro Kinase Assays-PLK1 immunoprecipitation-kinase assays were performed as described previously (27) and analyzed by immunoblotting using P-c-MYC (Ser 62 ) and c-MYC antibodies.
Quantitative Real Time PCR and PCR Array-Total RNA was prepared from cells using the Quick-RNA Mini Prep kit (Zymo Research), and cDNAs were generated using the iScript cDNA synthesis kit (Bio-Rad). SYBR Green reactions were done using a Bio-Rad CFX96 quantitative real time PCR system. For data analysis, raw counts were normalized to the housekeeping gene averaged for the same time point and condition (⌬C t ). Counts are reported as -fold change relative to the untreated control (2 Ϫ⌬⌬Ct ). All primers were designed and synthesized by Integrated DNA Technologies or Eurofins MWG Operon. Primers are listed in Table 1. Array samples were prepared according to the manufacturer's instructions by using RNeasy Plus Mini kits (Qiagen). Samples were analyzed using a human MYC target gene PCR array (Qiagen PAHS-177ZA). Genes were considered differentially expressed if they demonstrated a significant p value Յ0.05 and were at least 2-fold or greater up-regulated. An average of biological triplicates was used to generate heat maps by using Qiagen data analysis center software.
Chromatin Immunoprecipitation (ChIP) Assay-ChIP assays were done in 1-LN prostate cancer cells as described previously (33). ChIP enrichment for a primer set was evaluated by quantitative PCR as percentage of input and normalized to a negative primer set. FOSL1 and ID2 gene probes used for ChIP analysis were designed as described previously (33) and are listed in Table 2.
Statistical Analysis-Data are presented as mean Ϯ S.D. unless otherwise stated. A Student's t test was used to compare two groups for statistical significance analysis. p values Յ0.05 were considered as significant.

Results
␣ 2 M*-induced c-MYC Expression Promotes Tumor Growth in Soft Agar-We first determined the ability of ␣ 2 M* to promote proliferation by using an in vitro assay that measures anchorage-independent growth in soft agar. This assay demonstrates that ␣ 2 M* promotes a significant increase in colony formation in a variety of cancer cell lines (Fig. 1A). Furthermore, we analyzed the effect of ␣ 2 M* on cell proliferation in standard cell culture of the same panel of cancer cell lines. Consistent with our previous report, ␣ 2 M* significantly increased cell proliferation in these lines ( Fig. 1B) (9). This result confirms the potential role of ␣ 2 M* in cancer cell proliferation.
To determine the mechanism by which ␣ 2 M* is involved in cell proliferation, we examined the expression of c-MYC. ␣ 2 M* stimulation of 1-LN cells induces c-MYC protein expression in a dose-and time-dependent manner (Fig. 1C). Maximal c-MYC expression occurred with 100 pM ␣ 2 M* at 30 min. Of interest, the maximal effective dose of 100 pM is consistent with many previous studies, demonstrating that the maximal proliferative effect of ␣ 2 M* on proliferation occurs at this dose with a falloff in activity at 500 pM (for example, see Refs. 2 and 31). The reason for this dose effect is unknown; however, we used these conditions for subsequent experiments. Next we investigated whether ␣ 2 M* induces c-MYC at a post-translational or transcriptional level. 1-LN cells were treated with either cycloheximide or actinomycin D and then stimulated with ␣ 2 M*. ␣ 2 M* with either cycloheximide or actinomycin D decreased the c-MYC half-life, indicating that ␣ 2 M* induces c-MYC at the transcript level (Fig. 1D). ␣ 2 M* induced c-MYC transcription in a dose-and time-dependent manner ( Fig. 1E). At 500 pM ␣ 2 M*, c-MYC transcription continued to increase unlike the protein expression level. Recent reports indicate that c-MYC mRNA is targeted by several microRNAs to control its expression levels (34,35). Although actinomycin D decreased the ␣ 2 M*-induced change in c-MYC half-life, it is possible that ␣ 2 M* can induce the transcription of essential translational machineries or microRNAs required to modulate c-MYC translation.
To further substantiate the role of c-MYC activation by ␣ 2 M*, we suppressed c-MYC expression in 1-LN cells (Fig. 1F). ␣ 2 M* did not induce c-MYC transcript in these c-MYC silenced cells (Fig. 1G). Next we investigated the effect of ␣ 2 M* in Scr-sh-RNA and sh-c-MYC 1-LN cells by a soft agar assay. As expected, sh-c-MYC cells showed inhibition of colony formation, and ␣ 2 M* had no further effect (Fig. 1H). Indeed, RNA interference-mediated knockdown of c-MYC in 1-LN cells resulted in decreased proliferation, whereas ␣ 2 M* had no further effect (Fig. 1I). These data suggest that ␣ 2 M*-induced proliferation depends on c-MYC expression.
␣ 2 M* Signals through CS-GRP78 to Induce c-MYC Expression in a PDK1/PLK1 Signaling-dependent Manner-Our previous studies demonstrate that ␣ 2 M* induces signaling pathways through its receptor CS-GRP78 promoting cancer survival and metastasis (3,22,23). Recently, we produced a highly specific C38 mAb directed against the COOH-terminal domain of GRP78 that shows therapeutic efficacy in reducing tumor growth in vivo (15,32). We screened CS-GRP78 in multiple cancer cell lines ( Fig. 2A). To investigate whether ␣ 2 M* induces c-MYC expression through its receptor CS-GRP78, we treated the 1-LN cells with 50 g of C38 mAb for 6 h and then stimulated with ␣ 2 M*. We observed that targeting CS-GRP78 abrogated ␣ 2 M*-induced c-MYC transcription (Fig. 2B). Moreover, in a panel of cancer cell lines, C38 mAb treatment blocked ␣ 2 M*-mediated c-MYC protein expression (Fig. 2C, first lane). These findings demonstrate that ␣ 2 M* signaling through CS-GRP78 induces c-MYC expression.
To investigate the downstream kinases of CS-GRP78 action crucial for c-MYC induction, we investigated PDK1/PLK1 signaling, which induces MYC phosphorylation and protein accumulation (25,36). ␣ 2 M* induced strong phosphorylation of PDK1 (Ser 241 ) and increased phosphorylated PLK1 (Thr 210 ) in a dose-and time-dependent manner (data not shown). Further analyses in multiple cancer cell lines showed that targeting CS-GRP78 with C38 mAb abolished ␣ 2 M*-induced phosphorylation of PDK1 (Ser 241 ) and PLK1 (Thr 210 ) (Fig. 2C, third and fifth lanes). These data confirm that GRP78 is a functional molecular target on cancer cell surfaces.
In our previous studies, we identified the GRP78 primary amino acid sequence LIGRTWNDPSVQQDIKFL (Leu 98 -Leu 115 ) as the putative binding site for ␣ 2 M* (9). We next demonstrated the specificity of CS-GRP78 signaling by stimulating multiple cancer cell lines with ␣ 2 M* in the presence of WT (Leu 98 -Leu 115 ), mutant (K 113 -V), or scrambled GRP78 peptides. GRP78 WT peptide suppressed c-MYC induction and decreased ␣ 2 M*-dependent phosphorylation of both PDK1 and PLK1. In contrast, GRP78 mutant peptide did not affect ␣ 2 M*mediated c-MYC induction or phosphorylation of PDK1 and PLK1 (Fig. 2D). These results further demonstrate that ␣ 2 M* signals specifically through the GRP78 (Leu 98 -Leu 115 ) binding site to induce c-MYC expression.
␣ 2 M* Promotes PLK1 Interaction with c-MYC to Induce Phosphorylation of c-MYC in a PDK1-dependent Manner-To dissect the PDK1/PLK1 signaling pathway, we treated multiple cancer cell lines with the small molecule PDK1 inhibitor BX795 or PLK1 inhibitor BI2536 and then stimulated with ␣ 2 M*. Indeed, the PDK1 inhibitor BX795 abolished ␣ 2 M*-induced phosphorylation of PDK1 and expression of c-MYC at the pro- tein as well as transcriptional level (Fig. 3, A (first and fifth lanes) and B). Furthermore, like BX795, the PLK1 inhibitor BI2536 eliminated c-MYC protein expression and affected its transcription level but did not affect P-PDK1 (Fig. 3, A (third and fifth lanes) and B). These findings suggest that PLK1 functions as a downstream kinase to PDK1 in ␣ 2 M*/CS-GRP78-induced c-MYC expression.
To determine whether PDK1 is required for the PLK1 activation, we stimulated 1-LN cells with ␣ 2 M* followed by PDK1 immunoprecipitation. We demonstrated enhanced phosphorylation of endogenous PLK1 at Thr 210 that was abolished when cells were treated with BX795 or BI2536 (Fig. 3C). This suggests that PDK1 directly regulates PLK1 in cancer cells. Importantly, cells treated with C38 mAb also showed greatly reduced PLK1 phosphorylation. Together these data indicate that PDK1-induced PLK1 phosphorylation requires ␣ 2 M*/CS-GRP78 signaling.
We next investigated whether PLK1 directly regulates c-MYC activation. Through co-immunoprecipitation assays, we showed that ␣ 2 M* promoted interaction between PLK1 and c-MYC that is blocked by BX795, BI2536, or C38 mAb (Fig. 3D). We explored whether PLK1 kinase activity is required for the c-MYC phosphorylation and accumulation. We used in vitro kinase assays using endogenous PLK1 pulled down by antibody from ␣ 2 M*-stimulated scrambled and c-MYC-silenced 1-LN cells. We observed a robust induction of Ser 62 phosphorylation of recombinant c-MYC that was blocked in the presence of BX795 or BI2536 (Fig. 3, E and F). Importantly, PLK1 kinasedependent c-MYC phosphorylation was strongly abolished in cells treated with C38 mAb (Fig. 3, E and F). Phosphorylation of recombinant c-MYC in c-MYC-silenced cells confirms the specificity of the recombinant in vitro kinase assay (Fig. 3F). Thus we not only showed direct phosphorylation of c-MYC by PLK1 but also found that PLK1 activity with respect to c-MYC is crucially dependent on PDK1. Together these data strongly support the hypothesis that ␣ 2 M*/CS-GRP78 signaling potentiates PDK1/PLK1/c-MYC signaling in cancer cells.
␣ 2 M*/CS-GRP78 Signaling Axis Phosphorylates Histone H3 Ser 10 in a c-MYC-dependent Manner-To further substantiate the role of c-MYC activation, we stimulated multiple cancer cell lines with ␣ 2 M* in the absence or presence of C38 mAb. In ␣ 2 M*-stimulated cells, we observed induction of c-MYC Ser 62 phosphorylation that was accompanied by a similar pattern of phosphorylation of histone H3S10, which is known for the transcriptional activation of the gene. We further showed that C38 mAb and GRP78 WT peptide blocked ␣ 2 M*-mediated phosphorylation of c-MYC and P-histone H3S10, whereas c-MYC-silenced cells and GRP78 mutant peptide did not have any further effect (Fig. 4, A, B,  and C). Notably, in multiple cancer cell lines, BX795 or BI2536 treatment also inhibited ␣ 2 M*-induced phosphorylation of c-MYC and P-histone H3S10 as well as c-MYC accumulation (Fig. 4D). This finding demonstrates that ␣ 2 M*/CS-GRP78-mediated c-MYC activity is required for histone H3S10 phosphorylation and suggests a possible role of transcriptional activation of c-MYC target genes.

Role of ␣ 2 M*/CS-GRP78 Signaling in c-MYC-dependent Gene
Regulation-We next sought to determine whether ␣ 2 M*-mediated c-MYC induction activates specific sets of target genes that are characteristic of c-MYC transformed cells. To achieve that, we performed human MYC target gene PCR array analysis of ␣ 2 M*-stimulated 1-LN cells in the presence or absence of C38 mAb. ␣ 2 M* significantly induced expression of c-MYC, FOSL1, and ID2 genes among those genes that were up-regulated (Fig. 5A). Importantly, C38 mAb suppressed up-regulation of these genes by ␣ 2 M*, demonstrating a role for CS-GRP78 activation in regulating the c-MYC target genes.
We next validated c-MYC target genes FOSL1 and ID2 by quantitative PCR analysis. ␣ 2 M* induced FOSL1 and ID2 transcript levels in a dose-and time-dependent manner (data not shown). Moreover, silencing of c-MYC strongly inhibited FOSL1 and ID2 expression, and ␣ 2 M* had no further effect (Fig.  5B). To confirm that ␣ 2 M* induces the c-MYC target genes FOSL1 and ID2 through its receptor CS-GRP78, we treated the 1-LN cells with 50 g of C38 mAb for 6 h and then stimulated with ␣ 2 M*. We observed that targeting CS-GRP78 abrogated the ␣ 2 M*-induced FOSL1 and ID2 transcription level (Fig. 5C). These results demonstrate that ␣ 2 M*/CS-GRP78 signaling is  To show the capability of PDK1/PLK1 signaling in regulation of c-MYC target genes, we treated the 1-LN cell line with BX795 or BI2536 inhibitor. Each compound inhibited induction of FOSL1 and ID2 transcripts by ␣ 2 M* (Fig. 5D). Together these findings show that ␣ 2 M*/CS-GRP78 signaling modulates PDK1/PLK1 signaling to evoke c-MYC target genes.
␣ 2 M*/CS-GRP78 Signaling Is Required for the Transcriptional Activation of FOSL1 and ID2 Genes-To demonstrate direct binding of c-MYC to the FOSL1 and ID2 regulatory sequences, we performed a ChIP assay using an antibody against c-MYC followed by RT-quantitative PCR analysis in 1-LN cells. As expected, ␣ 2 M*-stimulated 1-LN cells showed increased c-MYC association at the FOSL1 enhancer and the ID2 E-box region, whereas C38 mAb, BX795, or BI2536 abol-ished c-MYC binding and reduced its expression (Fig. 6A). To further confirm the specificity of c-MYC binding to FOSL1 and ID2, we used as a negative primer a downstream region within the first intron of FOSL1 and ID2. These results establish that ␣ 2 M*/CS-GRP78 signaling recruits c-MYC to E-box elements of the c-MYC-regulated genes FOSL1 and ID2 through its downstream kinase PDK1/PLK1 signaling, thus contributing to malignant progression.
Next we analyzed whether ␣ 2 M*/CS-GRP78 signaling contributes to H3S10 phosphorylation at the FOSL1 and ID2 enhancer region. An analysis of the H3S10 phosphorylation signal in ␣ 2 M*-stimulated 1-LN cells showed increased phosphorylation of H3S10 at the FOSL1 and ID2 enhancer region, whereas C38 mAb, BX795, or BI2536 inhibited the H3S10 phosphorylation signal and reduced the expression of FOSL1 and ID2 (Fig. 6B). These results demonstrate that ␣ 2 M*/CS-GRP78  MAY 13, 2016 • VOLUME 291 • NUMBER 20 regulates the transcriptional activation of FOSL1 and ID2 genes with an increase of H3S10 phosphorylation at the enhancer region. Because in c-MYC-silenced cells the transcription level of FOSL1 and ID2 decreased (Fig. 5B), we analyzed whether c-MYC association with the FOSL1 and ID2 enhancer is required for H3S10 phosphorylation. Quantitative ChIP analysis demonstrated that ␣ 2 M* did not induce H3S10 phosphorylation at the FOSL1 and ID2 enhancer in c-MYC-silenced cells (Fig. 6C). These data indicate that H3S10 phosphorylation at the FOSL1 and ID2 enhancer is dependent on c-MYC.

CS-GRP78 Signaling Promotes c-MYC Activation
␣ 2 M*/CS-GRP78-induced PDK1/PLK1/c-MYC Signaling Drives Cell Proliferation and Tumorigenesis-To address the functional role of ␣ 2 M*/CS-GRP78-mediated PDK1/PLK1/c-MYC signaling in cancer cells, we examined the transforming capacity of ␣ 2 M* in multiple cancer cell lines by using the soft agar assay. ␣ 2 M* potentiated increased colony numbers in a panel of cancer cell lines, whereas BX795 or BI2536 treatment significantly decreased colony number. This result further demonstrates that PDK1/PLK1 signaling is required for ␣ 2 M*induced transformation. Moreover, targeting CS-GRP78 by C38 mAb abolished ␣ 2 M*-induced colony growth (Fig. 7A). Consistent with our previous studies (9), ␣ 2 M* significantly increased cell proliferation of multiple cancer cell lines, and this was reduced by BX795 or BI2536 treatment (Fig. 7B). This finding further demonstrates the role of PDK1/PLK1 signaling in cell growth. Indeed, C38 mAb treatment also resulted in more effective growth inhibition. Together these data support the conclusion that ␣ 2 M*-induced transformation depends on CS-GRP78-mediated PDK1/PLK1 signaling in cell proliferation.
Next, we evaluated the role of ␣ 2 M*/CS-GRP78 in regulating PDK1/PLK1 signaling and cancer cell survival. As expected, BX795, BI2536, or C38 mAb induced massive apoptosis as evidenced by strong detection of poly(ADP-ribose) polymerase cleavage (Fig. 7C). As in our previous studies, targeting CS-GRP78 with C38 mAb induces significant apoptosis (23). Taken together these results further support the role of ␣ 2 M*/CS-GRP78 signaling in cell proliferation and malignant progression.

Discussion
This study defines a new role for the ␣ 2 M*/CS-GRP78 signaling axis as a c-MYC-dependent modifier of chromatin. On the basis of biochemical and functional evidence, we show that ␣ 2 M*/CS-GRP78-dependent PDK1/PLK1 signaling is required for the transcriptional activation of a subset of c-MYC target genes and cell proliferation.
Cell surface expression GRP78 and its ligation by ␣ 2 M* are clearly implicated in the development of metastatic prostate cancer (19). Furthermore, CS-GRP78 is differentially expressed in cancer cells and stressed endothelial cells, providing a potential opportunity for highly specific therapeutic intervention (7)(8)(9)(10)(11)(12)(13). We and others have shown that cell surface GRP78 through its interaction with ␣ 2 M* functions as an upstream regulator of PI3K/Akt signaling (3, 23, 37). Recently, our in vivo model suggested that targeting CS-GRP78 by C38 mAb blocks the Akt pathway to prolong the survival of ovarian cancer-bearing mice (15). By using GRP78 WT (Leu 98 -Leu 115 ) and mutant (K 113 -V) peptides, we further show that ␣ 2 M* induced PDK1/ PLK1/c-MYC signaling through the GRP78 primary amino acid sequence Leu 98 -Leu 115 , thereby demonstrating that this region is essential for signaling and transcriptional activation of c-MYC. Although ␣ 2 M*/CS-GRP78 signaling is associated with PI3K/Akt signaling in oncogenesis, our study uncovered another arm of signaling that routes to PDK1/PLK1/c-MYC signaling to activate c-MYC target genes and promote malignant progression. Consistent with previous findings that ␣ 2 M*/ CS-GRP78 induces PDK1 activation, we now further show that PDK1 induces PLK1 phosphorylation to maintain the cell growth in cancer cells (22,25). Importantly, the pathway we identified using a chemical and genetic approach with a GRP78 WT and mutant peptide and C38 mAb treatment shows that ␣ 2 M*/CS-GRP78 activates PDK1 to function as an upstream regulator of PLK1 for c-MYC induction. This is in contrast to a previous report suggesting that PDK1 induces c-MYC at a post-translational but not transcriptional level (25,36). Further studies are needed to determine the mechanism by which ␣ 2 M*/CS-GRP78 induces the c-MYC transcript level. Recent reports highlight that phosphorylation of c-MYC at Ser 62 enhances c-MYC activity through regulation of protein stability in Ras-expressing cells and prostate and breast cancer cells (25,30,38,39). Consistent with previous reports, we showed by immunoprecipitation experiments after stimulation with ␣ 2 M* that PLK1 forms a complex with c-MYC and phosphorylates Ser 62 (25). Moreover, we have demonstrated that C38 mAb reduces c-MYC protein stability by repressing Ser 62 phosphorylation and abrogates transcriptional activity of c-MYC. According to the current view, phosphorylation at the NH 2 -terminal domain of H3 is required to loosen the interaction between DNA and nucleosome and/or to generate a platform to recruit additional regulatory factors as described in the histone code hypothesis (41). It was shown previously that after treatment with growth factors H3 is phosphorylated rapidly at Ser 10 by MDK1/MSK2 (33,(42)(43)(44). Our results confirm these previous findings and emphasize that ␣ 2 M*/CS-GRP78 phos-phorylates histone H3S10 as a result of c-MYC activity and that this is required for the transcriptional activation of genes mediated by c-MYC. It has been estimated that about 11% of cellular genes contain a functional E-box with which c-MYC can associate on the genome (45). In agreement with these data, ChIP analysis in ␣ 2 M*-stimulated 1-LN prostate cancer cells revealed that c-MYC recruitment to chromatin requires CS-GRP78 signaling. As noted, c-MYC is necessary to phosphorylate histone H3S10 to activate transcription of its target genes FOSL1 and ID2. This is in agreement with previous findings that phosphorylation of H3S10 is necessary for transcriptional activation of FOSL1 and ID2 genes (33). C38 mAb suppressed c-MYC target gene expression that is dependent on ␣ 2 M* treatment, suggesting that ␣ 2 M*/CS-GRP78 participates in transcriptional activation of c-MYC-regulated genes.
C38 mAb strongly inhibited the formation of ␣ 2 M*-dependent colonies in soft agar. In addition, PDK1 or PLK1 inhibition FIGURE 6. ␣ 2 M*/CS-GRP78 signaling mediates H3S10 phosphorylation and contributes to FOSL1 and ID2 transcription in a c-MYC-dependent manner. A, chromatin samples were obtained from 1-LN prostate cancer cell line stimulated with ␣ 2 M* (100 pM) for 30 min in the absence or presence of BX795 (2.5 mol/liter) or BI2536 (10 nmol/liter) for 16 h or C38 mAb (50 g) for 6 h. ChIP assays were performed with c-MYC antibody. Immunoprecipitated DNA was analyzed by quantitative PCR with primers to amplify the FOSL1 enhancer and ID2 upstream c-MYC binding sites (Ϫ1750/Ϫ1460). Real time quantitative PCR measurements of the immunoprecipitated DNA of the corresponding gene regions are expressed in thousandths (%) of the DNA inputs and normalized to the negative (Ϫve) primers. Values are mean Ϯ S.D. for three replicate samples from one representative experiment (n ϭ 3). Error bars represent S.D. B, phosphorylation of H3S10 at the FOSL1 and ID2 upstream element is dependent on ␣ 2 M*/CS-GRP78 signaling. ChIP assays were performed with anti-P-H3S10 antibody in 1LN cells as indicated in A. C, c-MYC silencing of 1LN cells alters H3S10 phosphorylation at the FOSL1 and ID2 enhancer. ChIP assays were performed with anit-P-H3S10 antibody in the sh-c-MYC 1LN cells as indicated in A. UT, untreated. resulted in decreased colony number. Although our experiments do not exclude the possibility that other signaling pathways might cooperate with c-MYC, our data strongly suggest that ␣ 2 M*/CS-GRP78-dependent PDK1/PLK1/c-MYC signaling induces phosphorylation of histone H3S10 that is necessary to regulate key genes required for c-MYC-dependent cell proliferation. Moreover, the main characteristic of the ␣ 2 M*/CS-GRP78-induced proliferation is that it is able to induce tumor initiation and progression. This study therefore provides a molecular mechanism for ␣ 2 M*/CS-GRP78-mediated c-MYCassociated gene regulation and c-MYC-dependent proliferation. Because mutations that alter c-MYC expression are among the most common found in human and animal cancers (46), it is conceivable that inhibiting PLK1 association with c-MYC and/or targeting CS-GRP78 by C38 mAb might represent a method for the treatment of c-MYC-driven cancers in human.
An intriguing finding of this study is the identification of ␣ 2 M*/CS-GRP78-dependent kinase activation upstream to PDK1/PLK1/c-MYC signaling to regulate cell proliferation (Fig. 8). We provide evidence that ␣ 2 M*/CS-GRP78 signaling activates PDK1 to induce PLK1 phosphorylation for the c-MYC activity. We further show here that PLK1 can directly bind to c-MYC and promote transcriptional activation of FOSL1 and ID2 genes by phosphorylating histone H3S10. Regardless of whether or not PDK1/PLK1 signaling regulates c-MYC stability through a similar or distinct mechanism, the regulation of c-MYC by ␣ 2 M*/CS-GRP78 signaling immediately suggests a therapeutic approach targeting tumors that are driven by c-MYC. Indeed, our data show a preferential killing by small molecule inhibitors of PDK1 or PLK1, thereby targeting c-MYC in cancer cells. Given that a clinical inhibitor of c-MYC is not available and targeting PLK1 with small molecule inhibitors such as BI2536 had a modest effect in patients with solid tumors (46,47), this study suggests that therapeutic targeting of CS-GRP78 by C38 mAb may yield a more favorable therapeutic index in c-MYC-associated tumors.
In patients with prostate cancer, the levels of both native and ␣ 2 M* in serum decrease during disease progression, whereas CS-GRP78 is often found to be overexpressed in human cancer cells when compared with the normal cells (9,48). Notably, we showed previously that ␣ 2 M* increases GRP78 expression in cancer cells (40). Prostate cancer cells should readily bind ␣ 2 M* from serum, thus activating CS-GRP78 to promote metastasis with poor survival. This notion is consistent with a recent report from Mandelin et al. (19) showing in vivo that cell surface occurrence of ␣ 2 M* and GRP78 is clearly implicated in the development of prostate cancer bone metastasis. In particular, the pathway we identified using multiple approaches in different cancer cell lines validates its relevance in human cancers. Importantly, targeting CS-GRP78 by specific monoclonal antibodies inhibits tumor growth in murine xenograft models of various tumors (12,15,37). Collectively, CS-GRP78 has attracted much attention as a potential therapeutic target in cancer. We propose that c-MYC can be an alternative pharmacodynamics marker for the evaluation of C38 mAb under preclinical and clinical development.
Author Contributions-U. G. performed all of the studies contained in this manuscript. M. G-G. gave advice with respect to experimental design and supplied the peptides used in some of the experiments. S. V. P. and U. G. designed and interpreted the studies reported here. U. G. and S. V. P. wrote the paper. All authors have read and approved the final version of this manuscript.