Inhibition of histone deacetylase suppresses EGF signaling pathways by destabilizing EGFR mRNA in ER-negative human breast cancer cells
- First Online:
- Cite this article as:
- Zhou, Q., Shaw, P.G. & Davidson, N.E. Breast Cancer Res Treat (2009) 117: 443. doi:10.1007/s10549-008-0148-5
- 323 Views
Estrogen receptor alpha (ER)-negative human breast cancer cells frequently overexpress epidermal growth factor receptor (EGFR) and respond poorly to endocrine therapies. Our previous studies demonstrate that histone deacetylation plays a key role in ER gene silencing, and ER expression can be restored with histone deacetylase (HDAC) inhibitors in ER-negative human breast cancer cells. Whether inhibition of HDAC also alters epidermal growth factor (EGF) signaling pathways is not defined. Here we present evidence that reexpression of ER protein by a clinically available HDAC inhibitor, suberoylanilide hydroxamic acid (SAHA or vorinostat), is coupled with loss of EGFR in ER-negative human breast cancer cells. Consistent with this observation, MDA-MB-231 cells, which are ER-negative and overexpress EGFR, that are engineered to express ER show a decrease in EGFR protein expression. Down-regulation of EGFR by SAHA results from attenuation of its mRNA stability. We also confirm that new protein synthesis is required for maintaining EGFR mRNA stability. Further experiments indicate that a decrease in EGFR abolished EGF-initiated signaling pathways including phosphorylated PAK1, p38MAPK and AKT. Thus, SAHA may not only reactivate silenced ER, but also simultaneously deplete EGFR expression. These data suggest that inhibition of HDAC is a promising epigenetic therapy for ER-negative human breast cancer.
Several studies have suggested that overexpression of the epidermal growth factor receptor (EGFR) is a marker for poor prognosis in breast cancer patients and is significantly correlated with the loss of endocrine sensitivity [1–4]. Upon binding to extracellular growth factors, EGFR (ErbB1/HER1), a receptor tyrosine kinase, undergoes receptor dimerization, kinase activation and transphosphorylation on its C-terminal regulatory tyrosine domain . The phosphorylated EGFR subsequently activates p21-activated kinase 1 (PAK1), extracellular signal-regulated kinase (ERK) mitogen-activated protein kinases (MAPK) and phosphoinositide 3-kinase (PI3K) signaling pathways, which contribute to the progression, invasion, and maintenance of the malignant phenotype in human cancers including breast cancer [1, 6–8]. It has been noted that the therapeutic benefit of endocrine therapy in hormone-dependent breast cancer is attenuated after activation of the EGFR pathway, principally via the PAK1 and MAPK signaling cascades [8–10].
There is a significant inverse association between expression levels of estrogen receptor α (ER) and EGFR in human breast cancer specimens . Our early work documented that this inverse association also exists in human breast cancer cell lines where it was shown that total EGFR mRNA levels correlate with the number of receptor-binding sites per cells and that transcriptional regulation is a key regulatory mechanism controlling EGFR expression in these cells . This finding suggests that these cell lines can serve as a model system to explore regulation of ER and EGFR in human breast cancers. Further, the potential for regulatory cross-talk between the ER and EGFR pathways has been extensively explored. It has been demonstrated that only ER-negative cells with overexpression of EGFR have DNase I hypersensitive sites in the first intron of EGFR [12, 13] and that certain transcription factors interact within this region to enhance transcriptional up-regulation of the EGFR in ER-negative breast cancer cells . These cells utilize the up-regulation of EGFR and activation of EGF signaling pathways to bypass the requirement of estrogen for cell proliferation that characterizes many ER-expressing human breast cancer cells. It appears that EGFR does not alter ER gene expression, but EGFR mRNA expression is negatively regulated by the presence of ER . Indeed constitutive expression of ER by transfection of ER cDNA into the ER-negative human breast cancer cells suppresses EGFR mRNA levels . Here it has been shown that ER represses EGFR through an interaction within a 96 bp region of a negative regulatory element that is located in the first intron of the EGFR gene [14, 15]. Thus, ER could serve as an internal restraint to oppose EGFR signaling in ER-positive breast cancer, whereas unopposed EGFR expression can promote progression of ER-negative breast cancer tumors [16, 17]. These studies imply that reintroduction of functional ER by pharmacological or genetic approaches may restore a brake to inhibit EGFR-initiated cell survival signaling in ER-negative breast cancer.
We and others have previously reported that ER can be epigenetically silenced as one possible mechanism that contributes to the ER negative phenotype in human breast cells [18, 19]. Transcriptional silencing of ER results from a close interplay between DNA methylation and histone modifications [20–23]. Histone modifications including acetylation and methylation of lysine residues on the amino-terminal tail have been defined as epigenetic modifiers as they create binding platforms for chromatin-associated proteins that silence gene transcription [21, 23]. HDACs are recruited to initiate stable and long-term epigenetic silencing of genes like ER in human breast cancer . HDAC inhibitors alone or in conjunction with DNA methyltransferase inhibitors can activate ER gene expression and restore tamoxifen sensitivity in ER-negative human breast cancer cells [22, 23], suggesting that the combination of epigenetic modifiers and antiestrogens may be a clinically relevant strategy for endocrine therapy. A number of small-molecule inhibitors have been developed to target HDACs, and are in various stages of clinical testing . Here we have used one such inhibitor, vorinostat or suberoylanilide hydroxamic acid (SAHA), to show that reexpression of ER by SAHA is accompanied by loss of EGFR expression, implying that its antineoplastic effects may not only include reactivation of ER, but also negative regulation of EGFR expression in ER-negative human breast cancer cells. Mechanistic studies have shown that down regulation of EGFR protein is associated with destabilization of EGFR mRNA, a process that requires new protein synthesis. Additionally, EGF-initiated EGFR signaling pathways, including PAK1, p38MAPK and AKT, are inhibited by SAHA treatment. Thus, our results suggest a novel inhibition of EGFR signaling pathways by SAHA in ER-negative human breast cancer cells.
Materials and methods
Cell culture conditions and reagents
MCF-7, MDA-MB-231, and MDA-MB-468 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% fetal bovine serum (Gemini BioProducts, Woodland, CA), and 2 μM l-glutamine (Invitrogen) at 37°C in a humidified CO2 incubator. SAHA was purchased from Biovision (Mountain View, CA). Antibodies against ERα (ER) and EGFR were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-PAK1, -AKT and -p38MAPK antibodies were from Cell Signaling Technology (Beverly, MA).
Plasmids and transfections
The ERα construct was described by Keen et al. . The Myc-tagged full-length PAK1 (Myc-PAK1) construct was a gift from Dr. Jonathan Chernoff (Fox Chase Cancer Center, Philadelphia, PA). For transient transfection, cells were plated at a density of 1 × 106 cells per 10 cm culture dish approximately 24 h before transfection. Plasmid was transfected into the cells with Oligofectamine 2000 (Invitrogen) as previously described . After 24 h of incubation, cells were then treated with or without the designated drugs.
RNA isolation, RT-PCR and real time PCR
RNA was harvested from cells using the TRIzol reagent (Invitrogen, Carlsbad, CA). cDNA was synthesized from 3 μg total RNA using MMLV reverse transcriptase (Invitrogen) and oligo dT primers (Invitrogen) at 37°C for 1 h . Conventional PCR was performed using cDNA samples. All reactions involved an initial denaturation at 95°C for 1 min, annealing at 55°C for 1 min and polymerization at 72°C for 2 min with final extension at 72°C for 5 min for a total of 30 cycles. The primers used for PCR were as follows: EGFR (forward primer, 5′-GAGAGGAGAACTGCCAGAA-3′; reverse primer, 5′-GTAGCATTTATGGAGAGTG-3′); actin (forward primer, 5′-ACC ATG GAT GAT GAT ATC GC-3′; reverse primer, 5′-ACA TGG CTG GGG TGT TGA AG-3′). PCR products were resolved by 1.5% agarose gel electrophoresis and visualized by ethidium bromide staining. Real time PCR was performed with cDNA samples using the ABI Prism 7900 Sequence Detection System (Applied Biosystems, Foster City, CA). Primers were as follows: EGFR (forward primer, 5′-GTCTGCCATGCCTTGTGCTC-3′; reverse primer, 5′-CTTGTCCACGCATTCCCT GC-3′); GAPDH (forward primer, 5′-GAA GGT GAA GGT CGG AGT C-3′; reverse primer, 5′-GAA GAT GGT GAT GGG ATT TC-3′). The data were normalized by the GAPDH housekeeping gene detection .
Cells were fixed with 1.5% formaldehyde, permeabilized with 0.5% Triton X-100, and blocked with 1% bovine serum albumin (BSA). Primary antibodies were added to blocking solution (1% BSA and 0.1% Triton X-100) for 1 h. Cells were washed, and secondary antibodies were added at 1:400 dilution. Slides were counterstained with DAPI. Images were viewed using the 20× objective lens with a Nikon confocal microscope.
Whole cell lysates were prepared by lysing the cells with 1% SDS and 10 mM Tris-HCl (pH 7.4). The supernatants were collected with microcentrifugation (14,000 rpm, 10 min) at 4°C. Equal amounts of protein were denatured in 4× Laemmeli’s sample buffer and separated on 10% polyacrylamide gels (GeneMate, ISC Biotechnology, Kaysville, UT). Separated proteins were transferred to PVDF membrane (BioRad, Hercules, CA). Membranes were blocked with 5% milk in TBS-Tween 20 overnight at 4°C with constant shaking and then probed with appropriate primary antibodies. Proteins were detected by chemiluminescence (ECL, Amersham, Piscataway, NJ). The expression of β-actin was used as a control.
Measurement of EGFR mRNA stability
Cells were treated with SAHA for 4 h and then blocked with 5 μg/ml actinomycin D. RNA was isolated after 0, 1, 2 and 5 h of actinomycin D treatment and EGFR mRNA levels were measured by real-time PCR analysis as described above. Stability of EGFR mRNA in the presence or absence of SAHA was estimated by plotting EGFR mRNA levels as a function of time. To test the role of new protein synthesis in SAHA-mediated effects on EGFR mRNA, cells were treated with 10 μg/ml cycloheximide for 2 h followed by SAHA for 4 h with actinomycin D for 6 h. The cells were then harvested for EGFR mRNA measurement.
Data are presented as mean ± sd. One way ANOVA (analysis of variance) followed by Bonferroni’s t test was used to assess statistically significantly differences between drug treated cells and vehicle control. Statistical analysis was done using GraphPad Prism for Windows version 4.00. P < 0.05 was considered statistically significant.
Reexpression of ER by SAHA is associated with concomitant reduction of EGFR
We next examined the effect of SAHA on EGFR expression. Time course studies show that 10 μM SAHA reduced EGFR protein expression within 12 h and this effect persisted for at least 24 h (Fig. 1b). As shown in Fig. 1b, treatment of MDA-MB-231 cells with SAHA (5–15 μM) for 24 h dramatically suppressed EGFR protein expression in a dose-dependent fashion as compared with untreated cells. Similar results were observed in SAHA-treated MDA-MB-468 cells.
Since SAHA-related reexpression of ER is coupled with loss of EGFR (Fig. 1a, b), we asked whether the decrease in EGFR expression might be linked to reexpression of ER. To test this possibility, ER-negative MDA-MB-231 cells were transiently transfected with an ER expression vector for 24 h. The transfected cells were harvested and whole cell lysates were analyzed by Western blotting for ER and EGFR expression. EGFR protein expression was undetectable in the ER-positive MCF-7 cells, but easily detected in ER-negative MDA-MB-231 cells (Fig. 1c). ER expression was appreciated in ER-transfected MDA-MB-231 cells and EGFR expression was reduced as compared with an empty vector control. Similar results were obtained in MDA-MB-231 cells that were stably transfected with the ER construct (data not shown). Collectively, these data suggest that reexpression of ER by pharmacological or genetic approaches in cells in which the ER is not normally expressed can suppress EGFR expression.
SAHA inhibits EGFR mRNA
SAHA inhibits EGF-initiated signaling pathway
Since EGF is a potent inducer of PAK1, MAPK and AKT pathways, it is necessary to explore if down-regulation of EGFR by inhibition of HDAC could potentially suppress these kinases. MDA-MB-231 and MDA-MB-468 cells were grown in phenol red-free medium supplemented with charcoal-stripped serum and exposed to EGF for 20 min. As expected EGF stimulated phosphorylation of PAK1 after 20 min in both cell lines without altering PAK1 protein expression (Fig. 3c). In contrast, EGF-induced phosphorylation of PAK1 was abolished by pretreatment with SAHA for 24 h. Figure 3c also shows that PAK1 protein was completely depleted by SAHA alone or cotreatment with EGF. These data support that depletion of EGFR and PAK1 by SAHA impairs the EGF/PAK1 signaling pathway. Accordingly, effects of SAHA on p38MAPK and AKT signaling were also examined. Robust phosphorylation of p38MAPK and AKT was observed upon exposure of either cell line to EGF, but was inhibited by pretreatment with SAHA. As shown in Fig. 3c, the protein levels of p38 MAPK and AKT were similar in untreated and treated cells. Together, these results suggest that depletion of EGFR by SAHA treatment leads to secondary inhibition of PAK1, MAPK and AKT signaling pathways in ER-negative human breast cancer cells, but by different mechanisms as SAHA leads to depletion of PAK1 protein and abolition of phosphorylation of p38MAPK and AKT.
SAHA suppresses PAK1 protein expression in the nucleus and cytoplasm
Activation of PAK1 by EGF is one of the major pathways that mediate or promote invasive phenotypes of breast cancer cells through actin cytoskeleton reorganization [27, 28]. Given the important role of PAK1 kinase in tumor cell migration and control of downstream signaling, it is necessary to further confirm SAHA-induced down-regulation of PAK1 protein, which has been shown in Fig. 3c.
To assess the effect of SAHA on the subcellular distribution of PAK1, MDA-MB-231 cells were transfected with Myc-tagged PAK1 and visualized by immunofluorescence microscopy. Figure 4c shows that overexpressed wild-type PAK1 is distributed throughout the cytoplasmic and nuclear compartments as observed in previous studies of breast cancer cells [29, 30]. After 24 h of SAHA treatment, these cells displayed decreased accumulation of PAK1 in both cytoplasmic and nuclear compartments.
About one-quarter of invasive breast cancers are characterized by a lack of ER expression. This estrogen independent phenotype is not only associated with the loss of ER, but is often also accompanied by high levels of EGFR . Compelling evidence indicates that histone deacetylation plays a key role in inactivation of ER gene expression [20–23]. As demonstrated by Chromatin Immunoprecipitation (ChIP) analysis, the ER CpG promoter is occupied by abundant HDAC1 and HDAC2 in ER-negative breast cancer cells [21, 23]. Recently, we reported that an HDAC inhibitor in early clinical trials, LBH589, could reactivate functional ER expression and enhance sensitivity to endocrine therapy in ER-negative human breast cancer cells , suggesting that inhibition of HDAC can epigenetically reactivate functional ER expression. Conversely, EGFR and intracellular signaling kinases, notably PAK1, MAPK and AKT, have been suggested to participate in the development of resistance for endocrine therapy [8–10]. The underlying mechanism of EGFR regulation of endocrine therapy involves direct phosphorylation of ER, leading to ligand-independent stimulation of its activation function-2 domain and the stimulation of estrogen-inducible genes . For example, active PAK1 signaling sustains S118 phosphorylation of ER, which may contribute toward the loss of antiestrogenic effect of tamoxifen . Therefore, the development of novel therapies that potentially target both ER and EGFR pathways will bear clinical significance.
In the work presented here, we demonstrate that inhibition of HDAC simultaneously modulates ER and EGFR expression in ER-negative human breast cancer cells. Two ER negative human cancer cell lines, MDA-MB-231 and MDA-MB-468, were used to confirm this conclusion. We observed that reexpression of ER by SAHA is coupled with the loss of EGFR mRNA stability in ER-negative breast cancer cells. Additionally, treatment with SAHA resulted in a decrease in EGF-initiated signaling pathways including PAK1, p38MAPK and AKT. Our data suggest that inhibition of HDAC may be useful against ER-negative breast cancer cells that express modest levels of EGFR.
An inverse relationship between ER and EGFR expression in human breast cancer cell lines has long been observed. MDA-MB-231 and MDA-MB-468 cells, which are ER-negative tumor cells, express detectable levels of EGFR, whereas MCF-7 cells, which are ER-positive breast cancer cells, show undetectable levels of EGFR by Western blotting. SAHA reactivates ER expression with a concomitant reduction of EGFR protein expression. Consistent with these data, EGFR protein levels were diminished in MDA-MB-231 cells that were engineered to express ER by transient or stable transfection. These studies suggest that EGFR expression may be regulated at least in part through an ER-mediated control mechanism. Most importantly, the cycloheximide-actinomycin experiments suggest that new protein synthesis is crucial for SAHA-induced destabilization of EGFR mRNA. The fact that SAHA reactivates ER raises the possibility that reexpressed ER and other key EGFR mRNA-binding proteins may mediate EGFR mRNA stabilization.
PAK1 kinase activity can be stimulated by several upstream signaling pathways including EGFR, HER2 and IGF-1 growth factor receptors. Activated PAK1 in turn phosphorylates AKT and p38MAPK kinase pathways, leading to stimulation of tumor cell proliferation and abrogation of tamoxifen action in hormone-sensitive breast cancer cells [8, 29, 33]. Since stimulation of EGFR by its ligand, EGF, strongly induces activation of PAK1, MAPK and ATK pathways, we selected these tyrosine kinases to serve as important biomarkers of activated EGFR, and explore if down-regulation of EGFR by inhibitors alters the EGFR signaling. MDA-MB-231 and MDA-MB-468 cells were treated with EGF, SAHA or a combination of both compounds. The status of PAK1, MAPK and AKT phosphorylation as indicators of kinase activation was examined by Western blot using phosphospecific antibodies. Serum-starved cells showed induction of phospho-PAK1 when treated with the EGFR ligand, EGF. These data demonstrate that EGFR is present and active in both cell model systems. Importantly, the phosphorylation status of the downstream effectors, MAPK and AKT, was well correlated with PAK1 phosphorylation. Interestingly, total PAK1 protein was depleted in SAHA- but not in EGF alone treated cells, although total MAPK and ATK protein expression was fairly consistent. We observed that EGF-stimulated phosphorylation of PAK1 cascades can be completely abolished by SAHA treatment. This likely reflects SAHA’s ability to reduce EGFR and PAK1 protein expression, thereby crippling EGF signaling pathways.
The significance of PAK1 activity has been highlighted in cell cycle regulation, mitosis and microtubule-organization in cytoplasm as well as in nucleus [28, 34, 35]. Immunohistochemical staining showed increased levels of PAK1 in the nucleus and cytoplasm in breast cancers from hormone resistant patients [29, 30, 36]. These findings have been confirmed and reinforced by our confocal analysis. We further verified that activity of PAK1 is stimulated by EGF. MDA-MB-231 cells exhibit a dramatic time-dependent increase in the amount of tyrosine-phosphorylated PAK1 without a change in total PAK1 protein over a 25-min time course. However, SAHA treatment abolishes protein levels of PAK1 in a time and dose-dependent manner. In addition to these observations, our confocal analysis demonstrates that SAHA inhibits distribution of exogenous PAK1 in the nucleus and cytoplasm in PAK1-Myc transfected cells. Once again, these results reflect an inhibitory effect of SAHA on PAK1 protein and also reveal a close correlation between EGFR and PAK1 protein levels. One possible explanation for these data is that depletion of EGFR by SAHA may cause PAK1 destabilization. This interpretation is supported by evidence that stimulation of EGFR enhances the level of EGFR-associated PAK1 in HcCaT cells . Further studies will address how inhibition of HDAC alters the interaction of EGFR with PAK1 or other adapter proteins such as Grb2. This study could possibly provide a deeper insight into the EGFR-PAK1 signaling pathway.
In conclusion, SAHA treatment results not only in reexpression of ER, but also in inhibition of EGFR expression in ER-negative human breast cancer cells. Mechanistically, reduction of EGFR protein observed upon treatment of ER-negative human breast cancer cells with SAHA is mediated by disruption of its mRNA stability. This effect is essential for the disruption of downstream EGF signaling including phosphorylated PAK1, MAPK and AKT. Our studies, together with these findings from others, suggest then that SAHA could be a clinically applicable treatment in ER-negative human breast cancer cells. The utility of this approach in the lab and the clinic is under evaluation.
This work was supported by grants from the National Institutes of Health (CA88843), Department of Defense (DAMD 17-03-1-0376), the Avon Foundation and FAMRI (YCSA-072084).