Hyaluronan-CD44 interaction promotes c-Jun signaling and miRNA21 expression leading to Bcl-2 expression and chemoresistance in breast cancer cells

MicroRNA-21 (miR-21) is associated with the development of solid tumors progression including breast cancer. In this study we investigated matrix hyaluronan (HA)-CD44 (a primary HA receptor) interaction with c-Jun N-Terminal Kinase (JNK)/c-Jun signaling in MDA-MB-468 breast cancer cells [a triple-negative (estrogen receptor-negative/progesterone receptor-negative/HER2-negative) breast cancer cell line]. Our results indicated that HA binding to CD44 promotes c-Jun nuclear translocation and transcriptional activation. Further analyses revealed that miR-21 is regulated by an upstream promoter containing AP1 binding site(s), and chromatin immunoprecipitation (CHIP) assays demonstrated that stimulation of miR-21 expression by HA/CD44 interaction is c-Jun-dependent in these breast cancer cells. This process results in an increase of the anti-apoptosis protein Bcl-2 and upregulation of inhibitors of the apoptosis family of proteins (IAPs) as well as chemoresistance in MDA-MB-468 cells. Treatment with c-Jun specific small interfering RNAs effectively blocks HA-mediated c-Jun signaling and abrogates miR-21 production as well as causes downregulation of survival proteins (Bcl-2 and IAPs) and enhancement of chemosensitivity. In addition, our results demonstrated that anti-miR-21 inhibitor not only downregulates Bcl-2/IAP expression but also increases chemosensitivity in HA-treated breast cancer cells. Together, these findings suggest that the HA/CD44-induced c-Jun signaling plays a pivotal role in miR-21 production leading to survival protein (Bcl-2/IAP) upregulation and chemoresistance in triple negative breast cancer cells such as MDA-MB-468 cell line. This novel HA/CD44-mediated c-Jun signaling pathway and miR-21 production provide a new drug target for the future intervention strategies to treat breast cancer.


Introduction
Matrix Hyaluronan (HA) is an anionic, nonsulfated glycosaminoglycan distributed widely throughout connective, epithelial, and neural tissues [1]. As a major component in the extracellular matrix of most mammalian tissues, HA contributes significantly to cell adhesion, proliferation and migration/invasion [2][3][4]. There is also a great deal of evidence linking high level of HA production in human carcinomas to aggressive phenotypes and metastasis, including the progression of breast cancer [2][3][4][5][6][7].
CD44 is a family of cell-surface glycoproteins that are expressed in a variety of tissues, including breast cancer tissues [2,3]. RHAMM whose cell surface form is now designated as CD168, was also found in breast cancer cells [8,9]. Both CD44 and RHAMM mediate hyaluronan signaling [10]. However, these two HA receptors likely regulate cellular signaling by different mechanisms because they are not homologous proteins, are compartmentalized differently in the cell, and differ in the way by which they bind to HA [10]. Since CD44 was identified as the first integral HA binding "receptor", HAmediated CD44 signaling has received a great deal of attention in cancer field. Both CD44 and HA are overexpressed/elevated at sites of tumor attachment [1,4]. HA binding to CD44 not only affects cell adhesion to extracellular matrix (ECM) components, but also stimulates a variety of tumor cell-specific functions leading to breast cancer progression [2,3,[11][12][13][14]. However, the oncogenic mechanism(s) occurring during HA-activated and CD44-specific breast cancer progression remain(s) to be determined.
Jun N-terminal kinases (JNKs) belong to the mitogenactivated protein kinase family, and are responsive to stress stimuli, such as cytokines, ultraviolet irradiation, heat shock, and osmotic shock [15]. Activation of JNKs by targeting phosphorylation of downstream effector proteins (e.g., c-Jun, ATF2, ELK1, SMAD4, p53 and HSF1) leads to a number of important cellular functions including cell growth, differentiation, survival and apoptosis [15,16]. Among these JNK-regulated target proteins, c-Jun was initially identified as the c-Fos-binding protein. The association between c-Jun and c-Fos forms the AP-1 early response transcription factor complex which then binds to DNA sequences located in the promoter regions of genes stimulated by externally added agonists [17]. In human cancer, the level of c-Jun and cfos mRNA and AP-1 expression has been shown to be elevated in drug-resistant tumor cells (such as etoposide resistant human leukemia K562 cells) as compared to the c-Jun/c-fos mRNA/AP-1 levels found in drug-sensitive parental lines [18]. Mitogenic stimulation of breast tumor cells (MCF-7 cell line) by insulin or insulin-like growth factors (IGFs) has been shown to promote c-Jun or c-fos upregulation and AP-1 activity [19]. Previous studies showed that persistent expression of c-Jun protein prevents stromal cells from entering apoptosis during the late secretory phase [20]. CD44 ligation blocks cell cycle progression of myeloid leukemia cells by downregulating c-Jun expression [20]. These observations suggest that c-Jun signaling is involved in regulating tumor cell growth, survival/anti-apoptosis and chemoresisitance.
MicroRNAs (miRNAs) are single-stranded RNAs of 21-25 nucleotides in length, which have been found to modulate gene expression at the posttranslational level [21]. MicroRNAs (miRNAs) are essential for normal development as modulators of gene expression. An estimated 30%-60% of the genome is regulated by miRNAmediated silencing [22], however aberrant expression of miRNAs is associated with many diseases, including cancer. Recent studies indicate that that some microRNAs upregulate the expression of its target gene by binding to the 3′ UTR [23,24]. Overexpression of miR-21 influences cell proliferation, invasion, metastasis and chemoresistance in different cancer cells including breast cancer cells [25][26][27]. The identified targets of miR-21 in human cancer cells include a tumor suppressor protein [Program Cell Death 4 (PDCD4)] [28]. A previous study indicated that HA-CD44 interaction promotes miR-21 production, and PDCD4 reduction in both breast cancer cells (MCF-7 cell line) and head and neck cancer cells (HSC-3 cell line) [25,29]. This event contributes to upregulation of inhibitors of apoptosis proteins (IAPs) and the multidrug resistant protein (MDR1)/P-glycoprotein (P-gp) resulting in anti-apoptosis and chemotherapy resistance in breast tumor cells (MCF-7 cell line) [25]. Thus, miR-21 is currently considered to be an oncogene. A recent report indicates that miR-21 can also stimulate the expression of an anti-apoptosis protein, Bcl-2 by binding directly to the 3′UTR of Bcl-2 mRNA [24]. Upregulation of Bcl-2 expression by miR-21 is associated with anti-apoptosis, chemoresistance and proliferation in pancreatic cancer cells [24]. The question of whether Bcl-2 expression is associated with miR-21 production in HA-treated breast tumor cells has not been addressed.
In this study we investigated a new HA/CD44-mediated c-Jun signaling pathway that regulates miR-21 production and chemoresistance in MDA-MB-468 cell line (a triple negative breast cancer cell line). Our results indicated that HA/CD44 activates c-Jun signaling which, in turn, stimulates miR-21 expression and function. These events lead to the production of an anti-apoptosis protein, Bcl-2 and upregulation of survival proteins (IAPs) and Doxorubicin chemoresistance in MDA-MB-468 cells. cells. Inhibition of c-Jun signaling or silencing miR-21 expression/function not only results in Bcl-2 downregulation, but also causes a reduction of survival protein expression and enhances chemosensitivity to Doxorubicin. Thus, our findings strongly support the contention that HA/ CD44-regulated c-Jun and miR-21 form a functional signaling axis that regulates tumor cell survival and Doxorubicin chemoresistance in triple negative breast cancer cells such as MDA-MB-468 cells.

HA-CD44 interaction activates JNK and c-Jun signaling in breast tumor cells
Previous studies [1,25,[30][31][32][33] indicated that HA/CD44mediated oncogenic signaling plays an important role in the development of several solid tumors including breast cancer. Among the signaling aberrations present in breast cancer, JNK and c-Jun signaling activation appears to be one of the critical pathways for the development of breast cancer [34,35]. Gene regulation by JNK-mediated c-Jun signaling generally requires specific phosphorylation of these two molecules [36,37]. Specifically, JNK phosphorylates c-Jun at Ser-63[pS63] residues within the transcriptional activation domain of c-Jun [36]. In this study we focused on the question of whether HA can regulate JNK activation and c-Jun signaling in breast tumor cells. To this end we examined a HA-mediated phosphorylation of JNK and c-Jun. Using anti-phospho-JNK[pS183]-and anti-phospho-c-Jun[pS63]-mediated immunoblot or anti-c-Jun immunoblot, respectively, we  [25,29,31,38]. In this study using immunofluorescence staining and confocal microscopy, we observed that both phosphorylated c-Jun[pS63] and c-Jun translocate from the cytosol to the nucleus after 30 min HA treatment ( Table 1) or without any HA treatment (Figure 2A and 2B). However, non-immune Rat IgG fails to reduce HA-mediated c-Jun and p-c-Jun nuclear accumulation (Table 1). These results indicate that nuclear translocation of c-Jun or pc-Jun is HA-dependent and CD44-specific. We also noted that neither phosphorylated-c-Jun[pS63] nor total c-Jun undergoes nuclear translocation in cells treated with JNK Inhibitor, 420116 plus HA (Figure 2A and Figure 2B).
The reason for showing both phospho-c-Jun and total c-Jun in MDA-MB-468 cells following HA treatment is to determine whether phosphorylated c-Jun represents the majority or a minority species of total c-Jun. The fact that the JNK inhibitor prevents nuclear translocation of both phospho-c-Jun and Jun suggests that majority of c-Jun is phosphorylated by JNK. This explains the effect of JNK inhibitor on blocking both phosphorylated c-Jun and total c-Jun nuclear accumulation in cells treated with HA. These findings strongly suggest that the HA-CD44 interaction promotes c-Jun nuclear translocation in MDA-MB-468 cells in a JNK-dependent manner.

Role of c-Jun in regulating miR-21 gene expression in HA/CD44
A previous study indicated that miR-21 is regulated by an upstream/enhancer promoter containing AP1 binding sites [39]. To examine whether c-Jun directly interacts    p-c-Jun association with the miR-21 promoter ( Figure 3Aa, b-lane 4 and 5). These findings suggest that the recruitment of c-Jun (also phospho-c-Jun) into the upstream/ enhancer region of miR-21 promoter site is HA-specific and CD44-dependent. To confirm the direct involvement of JNK-mediated c-Jun signaling in miR-21 gene upregulation, JNK activity was blocked by a JNK Inhibitor, 420116, ( Figure 3B) and c-Jun was downregulated by c-Jun small interfering RNA (siRNA), ( Figure 3C) followed by the miR-21 promoterspecific ChIP assay as described above. Our results indicate that (i) inhibition of c-JNK (but not vehicle control samples) ( Figure 3B Identical amplification products were detected in the positive controls from total input chromatin ( Figure 3A, B, Cd-all lines). Moreover, no amplification was seen in samples that were processed by IgG isotype control-mediated precipitation ( Figure 3A, B, C-c-all lanes). Therefore, we concluded that downregulation of JNK activity or c-Jun/ phospho-c-Jun expression by either JNK inhibitor (420116) or c-Jun siRNA is specific.
HA-CD44-activated JNK/c-Jun signaling stimulates miRNA-21 production in MDA-MB-468 Cells The expression of mature miR-21 is involved in breast cancer progression [31,40]. To determine whether miR-21 levels are increased following the binding of HA to CD44, we first prepared small RNAs followed by an RNase protection assay using the miRNA Detection Kit (Ambion). Our results indicated that the level of miR-21 is definitely increased in MDA-MB-468 cells treated Bcl-2 is identified as one of the target proteins induced by miR-21 [24]. Inhibitors of the apoptosis family of proteins (IAPs) [cIAP-1, cIAP-2 and X-linked IAPs (XIAP)] are frequently overexpressed by cancer cells. Importantly, upregulation of IAPs is linked to chemoresistance due to binding to caspases and suppressing apoptosis [41]. Here, we demonstrated that the expression of both Bcl2 and IAPs (cIAP1/cIAP2/XIAP) is greatly enhanced in cells treated with HA ( Figure 5A-a Finally, further analyses showed that the addition of HA enhances cell growth/survival and reduces apoptosis in untreated control cells or anti-CD44 antibody treated cells (but not non-immune rat IgG treated cells) (i.e., without chemotherapeutic drugs) and decreases the ability of Doxorubicin to induce tumor apoptosis and cell death (Table 2A). These observations indicated that HA causes both a decrease in apoptosis and an increase in breast tumor cell growth and survival (Table 2A & B) leading to the enhancement of chemoresistance (Table 2A) (Table 2A & B). Taken together, these findings strongly suggest that the HA/CD44-mediated JNK/c-Jun signaling pathways and miR-21 function represent new treatment targets to force tumor cells to undergo apoptosis/death and to overcome chemotherapy resistance in breast cancer cells.

Discussion
Hyaluronan (HA) is an important structural component of the extracellular matrix (ECM). In cancer patients, the level of HA is usually higher in malignant tumors than in corresponding benign or normal tissues, and in some tumor types the level of HA is predictive of malignancy [2][3][4][5][6][7]42]. In particular, HA level is elevated in the serum of breast cancer patients [4,42]. The aberrant HA  production by HA synthases [43][44][45] and HMW-HA degradation into LMW-HA by hyaluronidases [46] are thought to be closely associated with breast tumor cell progression [4]. HA binds specifically to CD44, a family of multifunctional transmembrane glycoproteins expressed in numerous cells and tissues, including breast tumor cells and various carcinoma tissues [2][3][4]. The crystal structure of the HA-CD44 complex was reported previously and a single HA binding site was identified [47]. CD44 is generally expressed in a variety of isoforms that are products of a single gene generated by alternative splicing of variant exons inserted into an extracellular membrane-proximal site [48,49]. CD44 is also expressed in tumor stem cells that have the unique ability to initiate tumor cell-specific properties [38,50]. In fact, CD44 is considered to be one of the important surface markers on cancer stem cells [38,50]. HA binding to CD44 is involved in the stimulation of both receptor kinases (e.g., ErbB2, EGFR and TGFβ receptors) and nonreceptor kinases (e.g., c-Src and ROK) [3] required for a variety of tumor cell-specific functions leading to tumor progression.
Abnormal JNK/c-Jun signaling also appears to play a critical role in oncogenesis [34,35]. JNK-activated c-Jun is a signal-transducing transcription factor of the AP-1 family that is implicated in cell cycle progression, differentiation and cell transformation [51]. It has a direct role in regulating the transcription of p53 and cyclinD1 [52,53]. It has also been shown that c-Jun accelerates  leukemogenesis and regulates the activation of genes required for cell cycle progression in tumor cells [51]. The AP-1 factor c-Jun is thought to act as a "bodyguard", preventing methylation of a distinct set of genes after oncogenic transformation [51]. Recently, c-Jun is found to trigger miR-21 transcription through AP-1 binding sites present in the miR-21 promotor region [39]. In this study we observed that HA-CD44 binding results in c-Jun (also causes phosphorylation of c-Jun) nuclear localization in MDA-MB-468 cells (Figure 2). Thus, identifying specific genes that are transcriptionally controlled by the JNK/c-Jun signaling during HA-CD44 interaction in the nucleus may be essential for understanding the disease mechanism occurring during breast cancer progression.
Overexpression of miR-21 is detected in various breast cancer cell lines and patient specimens [25,26]. Accumulating evidence indicates that miR-21 is closely associated with both cancer development and chemotherapy resistance [25]. The stem cell marker, Nanog, has been found to be involved in the regulation of pri-miRNA expression during cancer development [25]. Our previous work indicated that HA/CD44-activated PKCε promotes Nanog interaction with p68 and DROSHA leading to biosynthetic processing and production of miR-21 in breast tumor cells [25]. These findings suggest that HA/ CD44-mediated Nanog signaling is closely linked to miR-21 production and function during oncogenesis.
In this study, we provided new evidence that miR-21 expression is controlled by an upstream promoter/ enhancer containing AP-1 binding sites in MDA-MB-468 cells while chromatin immunoprecipitation (ChIP) assays demonstrate that stimulation of miR-21 production by HA is JNK and c-Jun-dependent in breast tumor cells (Figure 4). Most importantly, downregulation of JNK/c-Jun signaling (by treating cells with JNK inhibitor or c-Jun siRNA) or miR-21 (by treating cells with anti-miR-21 inhibitor) reduces the expression of the target protein, Bcl2, and anti-apoptotic proteins [e.g., IAPs (cIAP1/cIAP2/XIAP)] ( Figure 5) in breast tumor cells. Determining the cellular and molecular mechanisms involved in the regulation of these causal links between JNK/c-Jun signaling and miR-21 function, including Bcl2 and IAP upregulation, awaits further investigation.
Chemotherapy resistance is one of the primary causes of morbidity in patients diagnosed with solid tumors including breast cancer [54][55][56]. Chemotherapeutic agents, such as doxorubicin, are commonly used to inhibit DNA synthesis in the treatment of breast cancer patients [57]. In particular, the ability of doxorubicin to bind to DNA and/or produce free radicals is thought to be the mechanism for the induction of cytotoxic effects on tumor cells [57]. However, this drug often displays limited cytotoxic killing and anti-tumor effects due to chemoresistance which occurs in de novo tumor cells [57].
In this study we demonstrated that HA/CD44-activated JNK/c-Jun signaling and miR-21 increases survival protein, Bcl2, resulting in oncogenesis by enhancing the expression of inhibitors of anti-apoptosis proteins (IAPs) ( Figure 5). Furthermore, downregulation of HA/CD44activated JNK/c-Jun signaling (by JNK I nhibitor/c-Jun siRNA) and miR-21 production (by anti-miR-21 inhibitor) not only reduces Bcl2 upregulation ( Figure 5), but also inhibits the expression of survival proteins (e.g., c-IAP1, c-IAP2 and XIAP) ( Figure 5). Consequently, these signaling perturbation events contribute to apoptosis and chemosensitivity (Table 1). Furthermore, this newlydiscovered HA/CD44-activated JNK/c-Jun signaling pathway and miR-21 production/function should provide important new drug targets to cause tumor cell apoptosis and overcome chemotherapy resistance in breast tumor cells.
In summary (as shown in Figure 6), we propose that HA-CD44 binding (Step 1) promotes JNK and c-Jun (also serine phosphorylated JNK/c-Jun) (step 2). Subsequently, c-Jun (also p-c-Jun) translocates from the cytosol to the nucleus and interacts with an upstream/enhancer region (containing AP1 binding sites) of the miR-21 promoter (step 3), resulting in miR-21 gene expression (step 4) and mature miR-21 production (step 5). The resultant miR-21 then functions to upregulate the survival protein, Bcl2    In some cases, cell lysate of MDA-MB-468 cells (transfected with c-Jun siRNA or siRNA with scrambled sequences; or anti-miR-21 inhibitor or miRNA-negative control; or without any treatment) followed by HA (50 μg/ml) addition (or no HA or anti-CD44 antibody pretreatment plus HA) at 37°C were also immunoblotted using various immuno-reagents (e.g., mouse anti-Bcl-2 (2 μg/ml) or mouse anti-c-IAP-1 or mouse anti-c-IAP-2 antibody and anti-XIAP (2 μg/ml) or goat anti-actin (2 μg/ml) (as a loading control), respectively.

Chromatin immunoprecipitation assay
To examine whether c-Jun or phospho-c-Jun directly interacts with the upstream promoter/enhancer region (containing AP-1 binding site) of miR-21, chromatin immunoprecipitation (ChIP) assays was performed in MDA-MB-468 cells [pretreated with anti-CD44 antibody or JNK Inhibitor I, 420116 (20 μM)/vehicle control or transfected with c-Jun siRNA or siRNA with scrambled sequences] treated with HA (50 μg/ml) or without HA using a kit (EZ ChIP) from Millipore Corp according to the manufacturer's instructions. Crosslinked chromatin lysates were sonicated and diluted with ChIP sonication buffer plus protease inhibitors, divided and incubated with normal rabbit IgG or rabbit anti-c-Jun antibody or rabbit phospho-c-Jun[pS63] antibody at 4°C overnight, then precipitated with protein G agarose. Crosslinking was reversed by overnight at 65°C incubation; DNA fragments were then extracted with PCR purification kit, analyzed by PCR and quantitated by PCR using primer pairs specific for the miR-21 upstream promoter/enhancer region containing the c-Jun binding sites: forward primer: 5′-TGGATAAGGATGACGCACAG-3′ and reverse primer: 5′-TGGTTTGAACCAATTAATAAGGAAA-3′ on an agarose gel as described previously [25,29,38,60].

RNase protection assay analysis of mature miRNAs
Expression of miRNAs was qualitatively analyzed by RNase protection assay. For RNase protection assay, enriched small RNA isolated from MDA-MB-468 cells [transfected with scrambled sequence siRNA with (or without) anti-CD44 antibody or JNK Inhibitor I, 420116 (20 μM)/vehicle control or transfected with c-Jun siRNA or siRNA or anti-miR-21 inhibitor or miRNA-negative control in the presence or absence of HA for various time intervals (e.g., 0, 5 min, 10 min, 15 min, 30 min or 2 h) at 37°C] was enriched and purified using the mirVana miRNA Isolation kit (Ambion). RNA concentrations were verified by measuring absorbance (A 260 ) on the NanoDrop Spectrophotometer ND-1000 (NanoDrop). The mirVana miRNA probe construction kit (Ambion) was used to synthesize the 32 P-labeled miR-21 antisense probe and miR-191 probe loading control as described previously [25,29,38].

Immunofluorescence staining
MDA-MB-468 cells (untreated or pretreated with anti-CD44 antibody) were incubated with HA (50 μg/ml) at 37°C for 30 minutes or with no HA. These cells were then fixed with 2% paraformaldehyde. Subsequently, these cells were rendered permeable by ethanol treatment followed by incubating with Texas Red-conjugated anti-phospho-c-Jun[pS63] antibody or fluorescein (FITC)conjugated anti-c-Jun antibody followed by DAPI staining (a marker for nucleus). These fluorescence-labeled samples were then examined with a confocal laser scanning microscope.
Tumor cell growth and apoptosis assays MDA-MB-468 cells were either untreated or pretreated with anti-CD44 antibody or treated with JNK inhibitor, 420116 (20 μM) or transfected with c-Jun siRNA or siRNA with scrambled sequences or anti-miR-21 or miRNAnegative control] in the presence or absence of 50 μg/ml HA, as above. These cells were then plated in 96-well culture plates in 0.2 ml of Dulbecco's modified Eagle's medium/F12 medium supplement (GIBCO, Grand Island, NY) containing no serum for 24 h at 37°C in 5% CO 2 /95% air. These cells (5 × 10 3 cells/well) were then incubated with various concentrations of Doxorubicin (4 × 10 −9 M-1.75 × 10 −5 M) with no HA or with HA (50 μg/ml). After 24h incubation at 37°C, MTT-based growth assays were analyzed as described previously [25,30]. The percentage of absorbance relative to untreated controls (i.e., cells treated with neither HA nor chemotherapeutic drugs) was plotted as a linear function of drug concentration. The 50% inhibitory concentration (IC 50 ) was identified as a concentration of drug required to achieve a 50% growth inhibition relative to untreated controls. For apoptosis assay, FITC-conjugated Annexin V (for measuring apoptotic cells) using Apoptosis Detection Kit (Calbiochem, San Diego, CA) was used according to the manufacturer's protocol.