- Research Article
- Open Access
PDZK1 Is a Novel Factor in Breast Cancer That Is Indirectly Regulated by Estrogen through IGF-1R and Promotes Estrogen-Mediated Growth
Molecular Medicine volume 19, pages253–262(2013)
Although a relationship between PDZK1 expression and estrogen receptor (ER)-α stimulation has been suggested, the nature of such a connection and the function of PDZK1 in breast cancer remain unknown. Human tissue microarrays (cancer tissue: 262 cores; normal tissue: 87 cores) and breast cancer cell lines were used to conduct the study. We show that PDZK1 protein expression is tightly correlated with human breast malignancy, is negatively correlated with age and had no significant correlation with ER-α expression levels. PDZK1 exhibited an exclusive epithelial expression with mostly cytosolic subcellular localization. Additionally, 17β-estradiol induced PDZK1 expression above its basal level more than 24 h after treatment in MCF-7 cells. PDZK1 expression was indirectly regulated by ER-α stimulation, requiring insulinlike growth factor 1 receptor (IGF-1R) expression and function. The molecular link between PDZK1 and IGF-1R was supported by a significant correlation between protein and mRNA levels (r = 0.591, p < 0.001, and r = 0.537, p < 0.001, respectively) of the two factors in two different cohorts of human breast cancer tissues. Interestingly, PDZK1 knockdown in MCF-7 cells blocked ER-dependent growth and reduced c-Myc expression, whereas ectopic expression of PDZK1 enhanced cell proliferation in the presence or absence of 17β-estradiol potentially through an increase in c-Myc expression, suggesting that PDZK1 has oncogenic activity. PDKZ1 also appeared to interact with the Src/ER-α/epidermal growth factor receptor (EGFR) complex, but not with IGF-1R and enhanced EGFR-stimulated MEK/ERK1/2 signaling. Collectively, our results clarify the relationship between ER-α and PDZK1, propose a direct relationship between PDZK1 and IGF-1R, and identify a novel oncogenic activity for PDZK1 in breast cancer.
Breast cancer is the most common form of malignancy among women and is a leading cause of cancer-related deaths worldwide. It is the number one cause of cancer-related deaths in nonsmoking women in the US. Estrogen is a critical regulator of growth, differentiation and function in a wide array of target tissues, including the female and male reproductive tracts, the mammary gland and the cardiovascular and skeletal systems (1). Estrogen regulates both growth and differentiation of normal breast tissue; however, in pathological situations, the hormone participates in the initiation and progression of breast cancer (1). A great deal of work has demonstrated that the steroid hormone 17β-estradiol (E2) induces a signal essential for the transformation of normal cells to a malignant phenotype; however, the underlying mechanism(s) by which E2 participates in carcinogenesis is not well established (rev. in 2).
Estrogen receptor (ER) subtypes ER-α and ER-β mediate the action of E2 as well as those of a battery of natural and synthetic chemicals that share structural features with the hormone. The ER functions as a ligand-dependent regulator of transcription upon its direct recruitment to target genes or through interactions with other transcription factors (2). In addition, ER-α can participate in cytoplasmic and membrane-associated signaling processes when rapid physiological responses are required. Stimulation of the ER with E2 can induce a rapid signaling cascade that includes important signaling molecules, such as epidermal growth factor receptor (EGFR), Raf-1, mitogen-activated protein kinase (MAPK) and Akt, as well as several others. Such induction of important signaling processes ultimately contributes to the actions of E2 in the promotion of cancer cell proliferation and survival.
PDZK1, a 70-kDa adapter protein with four PDZ-interacting domains, which was first identified by Kocher et al. (3), is believed to regulate the levels of the scavenger receptor class B, type I (SR-BI) (or the scavenger receptor class B, member 1 [SCARB1]) in a posttran-scriptional manner (4). PDZK1 is found primarily at the apical brush-border membrane of the proximal tubules of the kidney (3). Lower levels can be detected in the liver, small intestine, pancreas, adrenal cortex, gastrointestinal tract and testis (3). What is well established is the fact that PDZ domain proteins are known to function as adaptor/scaffold proteins that influence multiple biological functions due to their interactions with a variety of plasma membrane transporters and receptors (5). Ghosh et al. (6) initially reported that PDZK1 can be an estrogen-responsive gene in breast cancer cells and suggested that it may be through an indirect mechanism. Such a relationship was recently strengthened by a clinical study establishing a significant association between plasma E2 levels and PDZK1 mRNA expression levels in ER-α–positive human breast cancer (7). These findings increased the potential importance of PDZK1 in breast cancer. However, the mechanism by which PDZK1 is induced upon E2 stimulation and the effects of such expression have yet to be elucidated. Accordingly, the present study aimed to examine the cellular and subcellular expression patterns of PDZK1 at the protein level in human breast cancer tissues, deciphering the mechanism by which ER-α stimulation leads to PDZK1 expression and defining the actual function of such protein.
Materials and Methods
Dulbecco’s modified Eagle medium (DMEM), penicillin, streptomycin and fetal bovine serum were purchased from Invitrogen/Life Technologies (Carlsbad, CA, USA). Charcoal/dextran-treated fetal bovine serum was purchased from Hyclone (Logan, UT, USA); E2, dihydrotestosterone (DHT), tamoxifen and cycloheximide were from Sigma-Aldrich (St. Louis, MO, USA); the synthetic androgen methyltrienolone (R1881) was from PerkinElmer (Waltham, MA, USA); U0126 and PP2 were from Promega (Madison, WI, USA); ICI 182,780 was from TOCRIS (Bristol, UK); and AG1024 was from Calbiochem (San Diego, CA, USA). AG1478 was from Cell Signaling Technology (Danvers, MA, USA). Unless otherwise indicated, all other drugs were purchased from Sigma-Aldrich.
Breast cancer and normal tissue microarray (TMA) sections (US Biomax, Rockville, MD, USA), which included a total of 262 neoplastic tissue samples and 87 normal breast adjacent tissue cores (Supplementary Table S1), were subjected to immunohistochemistry with antibodies to PDZK1 (ProteinTech Group, Chicago, IL, USA), insulinlike growth factor 1 receptor (IGF-1R) (Santa Cruz Biotechnology, Dallas, TX, USA) or ER-α (Novus Biologicals, Littleton, CO, USA) essentially as described previously (8). TMA slides were processed by using Biocare reagents in a Biocare Nemesis 7200 automated system (Biocare Medical, Concord, CA, USA). For negative controls, the entire immunohistochemistry method was performed on sections in the absence of the primary antibody. Stained slides were scanned using the ×40 objective with a Hamamatsu NanoZoomer (Hamamatsu, Bridgewater, NJ, USA). Immunoreactivity was analyzed by using Image-Pro Plus Version 6.0 software (Media Cybernetics, Rockville, MD, USA) as described previously (9,10). The measurement parameters included density mean, area sum and integrated optical density. The software system allows a computerized assessment of the density of the staining as a sum of the values for intensity of all the pixels of a counted region in an analyzed area as well as the total area in an unbiased manner. Threshold range of the colors of positive staining was selected in such a way that both faint and strong signals were detected without a high level of background. Density of immunoreactivity was then determined for all areas with a positive signal according to the weighted histoscore method (8):
In some situations, medium and high signals were combined and compared with percentage of cores with low or negative signals.
Cell Culture, Cell Proliferation, Transfection, Immunofluorescence Microscopy, Immunoblot Analysis and Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)
MCF-7, MDA-MB-231 and MCF-10A cells (all were obtained from ATCC, Manassas, VA, USA) were cultured according to ATCC specifications. These cell lines are authenticated by ATCC using short tandem repeat (STR) profiling. This PCR-based method allows the authentication of human cell lines with resolution down to the individual donor. Upon receipt from ATCC, the morphology was confirmed by microscopy and population-doubling times were determined by using the trypan blue dye exclusion method. Before treatment with E2, medium was changed to phenol red-free DMEM supplemented with 5% charcoal/dextran-treated fetal bovine serum (CDSS). Cell proliferation and DNA synthesis were measured by MTT assay and a 5-bromo-2-deoxyuridine (BrdU) assay kit (Roche, Indianapolis, IN, USA). MCF-7 cells were transiently transfected with specific siRNAs targeting ER-α (Santa Cruz Biotechnology), PDZK1 (AM16708 from Ambion/Life Technologies] or sc-106840 from Santa Cruz Biotechnology) or the negative control (scrambled) siRNA NEG3 (SA Bioscience, Frederick, MD, USA) by using Lipofectamine 2000 (Invitrogen/Life Technologies) according to the manufacturer’s instructions. Cells were also infected with a control and shRNA targeting the PDZK1 lentiviral vector (sc-108080 and sc-106849-V; Santa Cruz Biotechnology). The transduced cells were selected with puromycin dihydrochloride (Santa Cruz Biotechnology). Cells were then treated as described in the figure legends before being subjected to total RNA or protein preparation. Isolated RNA was reverse-transcribed, and the resulting cDNA was subjected to conventional PCR with primer sets (IDT, San Jose, CA, USA) specific to human PDZK1, IGF1-R, β-actin, or GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (Supplementary Table S2). The primer sequences for β-actin were selected to avoid amplification of pseudogenes as described by Raff et al. (9). Protein extracts were subjected to immunoblot analysis with antibodies to phospho-MEK, MEK, phospho(T202/Y204)-ERK1/2, ERK1/2, c-Src (32G6) and EGFR (all purchased from Cell Signaling Technology); ER-α, c-Myc or GADPH (all from Santa Cruz Biotechnology); or PDZK1(EPR3751) (Novus Biologicals). Immune complexes were detected with appropriate secondary antibodies and chemiluminescence reagents (Pierce, Rockford, IL, USA). Densitometry of immunoblot signals was quantified by using the VersaDoc™ imaging system (Bio-Rad Laboratories, Hercules, CA, USA). For immunofluorescence, cells were seeded on chamber slides and treated as described in the figure legends. Cells were then fixed, permeabilized and stained with antibodies to PDZK1 (Abcam, Cambridge, MA, USA) and with 4′-6-diamidino-2-phenylindole essentially as described (10). Cells were then examined with a Nikon fluorescence microscope.
Human Growth Factor Array
MCF-7 cells were treated with E2 for 6 h, after which protein extracts were collected. Changes in growth factor expression were analyzed by using human growth factor antibody arrays (RayBiotech, Hercules, CA, USA). Each array was incubated with 1 mL E2-treated or control cell lysates at 4°C overnight, and bound growth factors were detected by enhanced chemiluminescence according to the manufacturer’s instructions. The relative signal intensity was quantified by using the VersaDoc imaging system.
The lysates were precleared by incubation with 20 µL protein A/G plusagarose beads (Santa Cruz Biotechnology) for 1 h before an overnight incubation at 4°C with indicated specific antibodies or normal IgG. Protein A/G plus-agarose beads were then added to the lysates, and the mixtures were further incubated on a rotating wheel for 4 h at 4°C. The beads were pelleted and washed three times in wash buffer (0.1% NP-40, 50 mmol/L Tris, pH 8.0, 150 mmol/L NaCl, 2 µg/mL leupeptin, 2 µg/mL aprotinin, 1 mmol/L PMSF, 5 mmol/L NaF, 100 µmol/L Na3VO4). Antibody-antigen complexes bound to the beads were eluted in the sample buffer by boiling and were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (11).
Data are presented as means ± standard error of the mean (SEM) from at least three separate experiments. Comparisons between multiple groups were performed with one-way analysis of variance with Bonferroni test. Statistical significance was considered significant at p < 0.05. The Pearson correlation coefficient (r) was used to examine the correlation between different variables. All statistical summaries and analyses were performed by using GraphPad software, version 5 (La Jolla, CA, USA).
All supplementary materials are available online at https://doi.org/www.molmed.org .
Correlation Between PDZK1 Upregulation and Breast Malignancy
We initiated our studies by conducting a direct examination of PDZK1 expression in tissue microarray slides of both breast cancer and normal tissues. Clinical characteristics of the tissue microarray are displayed in Supplementary Table S1. Figures 1A and B show that PDZK1 expression was detectable in more than 60% of the carcinomas examined, suggesting a potentially high correlation with breast malignancy. PDZK1 immunoreactivity was largely absent or weak in both normal tissues and benign tumors. The expression of PDZK1 appeared to be exclusively epithelial, since no immunoreactivity was detected in stromal cells (Figure 1C). PDZK1 was detected in the cytoplasm, with occasional nuclear or perinuclear localization. PDZK1 immunoreactivity was significantly decreased in cancer tissues from women who were older than 55 years of age (Figure 1D). In fact, a significant negative correlation existed between PDZK1 immunoreactivity and the age of patients (r = −0.267, p = 0.002). This particular immunoreactivity remained considerably higher than that observed in normal or benign tissues. Surprisingly, however, there was no correlation between PDZK1 expression and ER-α levels (r = −0.105, p = 0.821). These results suggest a correlation between PDZK1 expression and breast carcinogenesis as well as provide a temporal window during which breast cancers would be positive for PDZK1. Interestingly, in ER-α–positive cancer tissues, PDZK1 expression was significantly higher in tissues from younger individuals (<40 years) compared with other ages (p < 0.001). In light of the recent report identifying a significant association between plasma E2 and PDZK1 mRNA levels in ER-α–positive breast cancers (7), our results suggest a more complex relationship between ER-α and PDZK1. Accordingly, it becomes imperative that the mechanism by which estrogen regulates PDZK1 expression is clarified and, more importantly, that the exact function of PDZK1 in breast cancer is explored.
PDZK1 is not an immediate early gene product of ER-α stimulation, but is critically required for E2-dependent growth of MCF-7 cells. Figure 2A shows that treatment of MCF-7 cells with 1 nmol/L E2 promoted a time-dependent increase in PDZK1 expression, as assessed by conventional PCR and immunoblot analysis. Notably, the accumulation of PDZK1 was rather slow, unlike ER-α–dependent genes such as GREB and pS2 (genes regulated by estrogen in breast cancer) that tend to exhibit much faster kinetics (6). Immunofluorescence analysis revealed that E2-induced PDZK1 exhibited a primarily cytosolic localization (Figure 2B) with some nuclear staining (Figure 2C). Such a pattern was similar to that observed in epithelial cells of breast cancer tissues (Figure 1C). Figure 2D shows that blockage of E2-mediated signaling by the ER-α antagonist fulvestrant (ICI182.780) or the selective estrogen modulator tamoxifen markedly blocked the induction of PDZK1 expression in response to the hormone, demonstrating the link between E2, ER-α and PDZK1. Surprisingly, in contrast to the reducing effect of the ER-α antagonists on E2-induced PDZK1 expression, inhibition of EGFR, a receptor that is rapidly activated upon ER-α stimulation (12), by its antagonist AG1478, did not exert any noticeable effect (Figure 2D). Additionally, treatment of the ER-α–deficient cell lines MDA-MB-231 and MCF-10A with E2 did not result in any detectable PDZK1 expression (Supplementary Figure S1), supporting the notion that E2-induced PDZK1 expression is specific for ER-α stimulation. Neither a natural (DHT) nor synthetic (R1881) androgen induced PDZK1 expression in MCF-7, MDA-MB-231 or MCF-10A cells (Supplementary Figure S1).
A critical outcome of ER-α signaling is its promotion of cellular growth. To determine whether a relationship exists between the E2-induced increase in PDZK1 expression and cell growth, we examined whether PDZK1 knockdown by siRNA would affect E2-promoted growth of MCF-7 cells. Treatment of MCF-7 cells with PDZK1-specific siRNAs, which markedly reduced the expression of the PDZK1 protein (Figure 2E, inset), coincided with the inhibition of E2-induced cell growth, as assessed by the MTT assay (Figure 2E) and confirmed by BrdU incorporation (Supplementary Figure S2). The approximate 40% increase in growth in response to E2 is typical and has been reported previously by others (13). The effect of PDZK1 knockdown on E2-promoted growth of MCF-7 cells was also confirmed by using a lentiviral vector encoding shRNA targeting PDZK1 (Supplementary Figure S3). Figure 2F shows that PDZK1 knockdown significantly reduced the number of cells in the S phase upon E2 treatment, suggesting a potential involvement of the protein in E2-induced entry into the S phase of the cell cycle. Such an effect on the cell cycle coincided with a marked decrease in E2-induced c-Myc expression (Figure 2G). The connection among ER-α, PDZK1 and c-Myc was subsequently confirmed by knocking down ER-α, which resulted in a major decrease in PDZK1 (Supplementary Figure S4). Such decrease coincided, as expected, with a reduction in c-Myc (Figure 2H). Together, these results strongly suggest that PDZK1 is not only an estrogen-responsive protein, but also plays a critical role in the induction process of cell growth upon ER-α stimulation by E2.
PDZK1 Expression Is Not Directly Controlled by E2 and Requires Expression of an Intermediate ER-α–Regulated Gene Product
As noted above, the kinetics of PDZK1 expression was relatively slow and not reflective of the rapid signal transduction normally induced by ER-α stimulation. This relationship was complicated even more by the lack of correlation between the two factors in human breast cancer tissues. Accordingly, it is conceivable that ER-α stimulation may be inducing an intermediate factor, which, in turn, induces PDZK1 expression. To test such a hypothesis, we examined the effects of inhibiting protein synthesis, a cycloheximide-sensitive event, on PDZK1 mRNA expression, a cycloheximide-resistant event. Figure 3A shows that E2-induced PDZK1 mRNA expression was almost completely blocked by cycloheximide, suggesting that E2 induces a protein or combination of proteins required for PDZK1 expression. These results confirm the observation made by Weigel’s group, who made the original observation on the dependence of PDZK1 on estrogen and actually suggested that the relationship between estrogen and PDZK1 might be indirect (6). To determine the critical time window during which this factor(s) was produced, we subjected cells to treatment with cycloheximide at different time intervals after E2 stimulation, as depicted in Figure 3B. Figure 3C shows that inhibition of protein synthesis 3–6 h after stimulation with E2 is a critical time window during which the unknown factor is synthesized. Later exposure to the drug could not effectively block PDZK1 expression upon E2 exposure.
IGF-1R Expression Is Critical for PDZK1 Expression on E2 Stimulation of MCF-7 Cells
To identify the potential factor that is responsible for PDZK1 expression, we subjected protein extracts from MCF-7 cells treated with E2 for 6 h to a growth factor array. Analysis of these results revealed that one of the most prominent candidates was IGF-1R (Figure 4A). Indeed, E2 induced IGF-1R expression above basal levels as early as 3 h after treatment in MCF-7 cells; IGF-1R reached maximum levels at approximately 12 h (Figure 4B), which remained high and sustained for at least 48 h after treatment. Figure 4C shows the dependence of IGF-1R expression above basal levels on ER-α stimulation, since knockdown of ER-α by siRNA drastically reduced expression of the receptor after E2-treatment in MCF-7 cells. Inhibition of ER-α activity by ICI182.780 or tamoxifen also markedly reduced IGF-1R expression in response to E2 treatment; however, treatment of cells with the IGF-1R antagonist AG1024 did not have any effect on E2-induced IGF-1R expression (Supplementary Figure S5).
To establish the involvement of IGF-1R in the E2-induced expression of PDZK1, we examined whether inhibition of IGF-1R-associated signaling by its potent inhibitor AG1024 would block PDZK1 expression in response to E2 treatment in MCF-7 cells. Further, we determined whether direct stimulation of IGF-1R would result in PDZK1 expression in the absence of E2. Figure 4E shows that AG1024 markedly reduced PDZK1 expression in a dose-dependent manner. More importantly, stimulation of IGF-1R with IGF-1 promoted rather fast expression of PDZK1; a noticeable expression level was observed after 3–6 h of treatment. Together, these results provide additional support for the involvement of IGF-1R in E2-induced expression of PDZK1.
Correlation Between IGF-1R and PDZK1 Expression in Human Breast Cancer at the Protein and mRNA Levels
We tested whether the relationship between PDZK1 and IGF-1R expression that was observed in the cell culture model was also present in human breast cancer tissues using tissue microarray slides. Analysis of the immunoreactivity to the two proteins in serial sections revealed a significant correlation (r = 0.591; p < 0.001), thus supporting the association between IGF1-R expression and PDZK1 upregulation (Figure 5A). Figure 5B depicts two examples of breast cancer tissue cores that are either positive or negative, which are respectively positive or negative for PDZK1. Using data deposited in the public domain Gene Expression Omnibus (GEO, GDS3097) Profiles database (14) and conducted by Boersma et al. (15), we examined the correlation between IGF1-R (203628_at) and PDZK1 (205380_at) mRNA levels in 48 fresh-frozen excised breast cancer tissues procured from a different cohort of patients. Figure 5C shows a significant correlation between the two transcripts (r = 0.537; p < 0.001), which lends additional independent support to the positive relationship between PDZK1 and IGF-1R in breast cancer.
Ectopic Expression of PDZK1 Stimulates Growth and Enhances E2-Promoted Growth of MCF-7 Cells
Given the correlation between PDZK1 expression and human breast cancer and the potential role of PDZK1 in ER-α–stimulated cell growth, we investigated the mechanism of action of PDZK1 in breast cancer cell proliferation. Initially, we addressed whether mere expression of PDZK1 was sufficient to promote growth of MCF-7 cells. To this end, MCF-7 cells were transiently transfected with an expression vector encoding human PDZK1 or GFP-expressing control vector (pcDNA3.1; Figure 6A), and cell growth was monitored in the absence of E2. Figure 6B shows that ectopic expression of PDZK1 significantly enhanced growth of MCF-7 cells (~35%; p < 0.005), as assessed by an MTT assay. Interestingly, the level of increase in cell growth was similar to that achieved by E2 treatment (Figure 2E). Given that PDZK1 knockdown reduced the E2-stimulated increase in c-Myc expression, we postulated that PDZK1 overexpression may consequently increase c-Myc in MCF-7 cells in the absence of E2. Indeed, ectopic expression of PDZK1 promoted a marked elevation in the level of c-Myc (Figure 6C). When growth of PDZK1-transfected MCF-7 cells was monitored in the presence of E2, proliferation of the cells was substantially enhanced (~80%; Figure 6D). The latter results were confirmed by direct cell counting with Trypan blue (Supplementary Figure S6). The substantial increase in the growth of MCF-7 cells transfected with the PDZK1-expressing vector did not appear to be associated with a greater elevation in PDZK1 upon E2 treatment, since the cumulative expression of PDZK1 was similar between untreated and E2-treated cells (Figure 6E). Together, these results clearly suggest an oncogenic function for PDZK1 in breast cancer cells and indicate that PDZK1 can enhance the growth-promoting activity of E2.
PDZK1 potentially stimulates growth of MCF-7 cells through interaction with the Src/ER-α/EGFR complex and enhancement of EGF-mediated signal transduction. Given PDZK1-mediated enhancement of cell growth, especially in response to ER-α stimulation, the question of how such protein influences E2-associated signaling events became important. It is noteworthy that PDZK1 was reported to interact with membrane receptors, including SR-BI, which appears to be essential for the initiation of its associated signal transduction in endothelium (16). c-Src is considered to be a critical factor that mediates signal transduction initiated by a number of receptor molecules including ER-α, IGF-1R and EGFR to ultimately promote cell growth (17). Inhibition of c-Src by PP2 was as efficient in inhibiting ERK1/2 phosphorylation in response to E2 stimulation as the MEK inhibitor U0126 (Supplementary Figure S7). For this purpose, we targeted c-Src to investigate whether PDZK1 influences E2-induced growth of MCF-7 cells by interacting with such key ER-α-associated receptor complex(es). Figure 7A demonstrates, as expected, an interaction among c-Src, ER-α and EGFR by using a coimmunoprecipitation assay. Interestingly, PDZK1 was also part of that complex in extracts from E2-treated cells, suggesting a potential interaction among the protein, c-Src and the receptors. c-Src and PDZK1 were not immunoprecipitated with control IgG (Supplementary Figure S8A). The interaction of PDZK1 and c-Src with IGF-1R was also tested but not detected (Supplementary Figure S8B), which prompted us to verify such lack of interaction by immunoprecipitation with antibodies against the receptor. Figure 7B shows that IGF-1R appeared not to bind to PDZK1 despite the large amount of the protein that was induced by E2 treatment.
In a recent comprehensive study, Santen’s group (18) reported that IGF-1R functions upstream of EGFR in a linear pathway, promoted as a result estrogen-mediated activation of MAPK and subsequent cell growth stimulation. Accordingly, we next wished to determine the consequence of the interaction of PDZK1 with the c-Src/ER-α/EGFR complex in EGFR-stimulated signal transduction. Figure 6C shows that stimulation of MCF-7 cells by EGF induced a fast and transient activation of ERK1/2. Such activation was greatly enhanced in cells that were treated with E2 for 48 h. PDZK1 knockdown by siRNA greatly dampened such ERK1/2 stimulation of E2-treated cells, strongly suggesting a causal relationship between expression of PDZK1 and enhancement of ERK1/2 activation. This enhancement of the ERK1/2 pathway may, in part, explain the involvement of PDZK1 in enhancing E2-mediated growth of MCF-7 cells. To define the exact underlying mechanism by which PDZK1 influences EGFR signal transduction and cell growth requires much more extensive and detailed studies.
The connection between estrogen and PDZK1 in breast cancer has recently received additional support from a clinical study conducted by Dunbier et al. (7), which showed a significant association between plasma E2 and PDZK1 mRNA levels in ER-α–positive breast cancers. Our results lend support for this association by providing the specific characteristics of protein distribution and its exclusive epithelial expression pattern with a primarily cytosolic and occasional nuclear localization in cancerous tissues, as assessed by immunohistochemistry. Interestingly, the expression profile of PDZK1 in human breast cancer appears to be prominent in younger women but decreased with age. Our findings become even more important in light of our mechanistic studies using an in vitro system with the estrogen-responsive breast cancer cell line MCF-7. Surprisingly, however, our study found no correlation between PDZK1 and ER-α expression levels, suggesting a more complex relationship. A more robust correlation was identified between PDZK1 and IGF-1R expression. The in vitro studies clarified this relationship and demonstrated that a potential link between E2/ER-α and PDZK1 expression may be IGF-1R, which functions as an ER-α–dependent intermediate factor that is required for PDZK1 expression. A critical finding of the current study was the determination of an actual function for PDZK1 as a growth-promoting factor in breast cancer cells. Such function is clearly different from its well-known scaffolding role in SR-BI protein stability and function in HDL signaling (5). The ability of PDZK1 to contribute to cell growth appears to be linked to c-Myc expression and enhancement of EGFR-associated signal transduction.
Our study shows that PDZK1 protein expression was undetectable in normal breast tissues or in benign tumors. In contrast, levels of PDZK1 protein is elevated in breast cancer tissues from women younger than 55 years of age. The levels of PDZK1 decrease in tissues from women older than 55 years of age. It is tempting to speculate that this result may be linked to the decreasing levels of plasma estrogen (19,20). This potential connection becomes even more important in light of the report that intratumoral estrogen levels could reach levels as high as 20 times that detected in the plasma of affected individuals (21). What renders the PDZK1 gene responsive to estrogen is also unclear. Our in vitro studies suggest that a potential requirement for PDZK1 expression is an increase in IGF-1R expression and activity. It is important to note that IGF-1R may not be the sole intermediate factor. Many epidemiological and mechanistic studies are necessary to identify the reason(s) behind the observed expression pattern of PDZK1 and whether these changes are critical determinants in breast carcinogenesis.
For more than a decade now, PDZK1 was believed to be a direct product of ER-α stimulation (22–24), although Weigel’s group suggested the potentially indirect relationship between estrogen and PDZK1 (6). Our results confirm the indirect relationship between PDZK1 and ER-α and showed that an intermediate E2-stimulated factor was required for PDZK1 induction. This factor was identified as IGF-1R. Indeed, direct stimulation of IGF-1R by IGF-1 in MCF-7 cells resulted in a fast and persistent expression of PDZK1 mRNA and protein. Additionally, expression of PDZK1 in E2-stimulated MCF-7 cells was completely blocked upon IGF-1R inhibition. The relationship between IGF-1R and PDZK1 was further supported by the strong correlation that existed between the proteins in human breast cancer tissues. This relationship was strengthened by the significant correlation between the mRNA levels of the two factors, as assessed in data deposited in the GEO Profiles database (14) and conducted by Boersma et al. (15). Additional support for the relationship between PDZK1 and IGF-1R in breast cancer comes from the finding that 100% of the samples tested in the latter study with high levels of PDZK1 mRNA exhibited moderate to high levels of IGF-1R. Conversely, 100% of the samples with low levels of IGF-1R mRNA exhibited low levels of PDZK1.
The correlation between plasma E2 and PDZK1 in ER-α–positive breast cancer tissues (r = 0.310, p = 0.0014) reported by Dunbier et al. (7) was lower than that between PDZK1 and IGF-1R (r = 0.559, p = 0.0001) observed in our study. Our assessment of ER-α was not limited to positivity but rather to protein levels, which we believe is more reflective of the potential response of breast cancer cells to estrogen. However, it is also possible that the intratumoral or plasma estrogen levels may influence the correlation (21). ER-α–positive breast cancer cells would not elicit an ER-α response if estrogen levels are low. It is important to acknowledge that the assessment of ER-α protein levels in tissues that come into contact with estrogen may be complicated by the fact that stimulation of ER-α by estrogen downregulates its receptor. This result was observed in MCF-7 cells (Supplementary Figure S4) and originally reported by Saceda et al. (25). Accordingly, the lack of correlation between ER-α and PDZK1 in the present study may be connected to such an event. Interestingly, MacKay et al. (23) recently showed that profound changes in PDZK1 gene expression in addition to other ER-α–responsive genes can be observed in ER-α–positive breast cancer tissues isolated from patients who were treated with aromatase inhibitors for 2 wks before surgery. A closer examination of the supplementary data presented in the latter study revealed a pronounced effect on IGF-1R expression levels, suggesting that the effect on PDZK1 by aromatase inhibitors may be associated with a reduction in IGF-1R. It is noteworthy that our results do not exclude a role for ER-α in PDZK1 expression but rather establish a stronger and direct link with IGF-1R and may explain why PDZK1 can be expressed in ER-α–negative breast cancers (6).
A remarkable feature of PDZK1 revealed by our study is that, although the expression of the protein is not directly regulated by ER-α stimulation, PDZK1 can influence the function of ER-α through its interaction with EGFR and Src. c-Src, a non-receptor tyrosine kinase, promotes the transduction of signaling receptors (such as IGF-1R, ER-α and EGFR) that regulate proliferation, survival, cell adhesion and migration (17,26). The involvement of c-Src in extra-nuclear signaling of ER-α is well established (26). The ability of PDZK1 to promote cell growth may be associated with the induction of c-Myc. Such a relationship was supported by the fact that knockdown of PDZK1 led to reduced c-Myc expression upon E2 treatment and ectopic PDZK1 overexpression correlated with an increase in c-Myc levels. The mechanism by which PDZK1 induces c-Myc expression is unclear. However, the association of PDZK1 with key signaling molecules, including EGFR, c-Src and ER-α, suggests that growth factor signaling is involved. Such a possibility is supported by the finding that E2-induced PDZK1 expression substantially increased ERK1/2 phosphorylation upon treatment with EGF. Given that ERK1/2 activation is required for c-Myc expression (27), the enhanced signal might be responsible for the involvement of PDZK1 in c-Myc expression. However, such a possibility remains to be determined with more detailed experimentation. Interestingly, Zhu et al. (28) reported that although PDZK1 is uniquely required for signaling by HDL through SR-BI, the interaction between c-Src and SR-BI is not influenced by PDZK1 in endothelial cells, suggesting a more complex role for the protein in breast cancer cells. Although EGFR signaling is enhanced by PDZK1, EGFR signaling does not appear to be involved in PDZK1 expression, since direct stimulation of cells with EGF did not result in PDZK1 expression and inhibition of EGFR activity by AG1478 did not suppress E2-stimulated PDZK1 expression (Figure 2).
Together, our findings provide further evidence for a relationship between E2, ER-α and PDZK1 in breast cancer cells and clarify the molecular mechanisms that underlie PDZK1 gene and protein expression. Our findings do not challenge the relationship between ER-α and PDZK1 but, instead, reveal an indirect association between these proteins. Interestingly, PDZK1 appears to have growth-promoting activity in the absence of ER-α stimulation. How PDZK1 promotes c-Myc expression without an external stimulus is puzzling to us but may be due to continuous interaction with c-Src and the promotion of persistent signaling, leading to expression of the oncogene and subsequent growth promotion. Much work remains to decipher the exact mechanism by which PDZK1 overexpression promotes growth of breast cancer cells. Taken together, our results suggest that PDZK1 may be a promising therapeutic target for inhibiting the growth of breast cancer cells.
The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
Bai Z, Gust R. (2009) Breast cancer, estrogen receptor and ligands. Arch. Pharm. (Weinheim). 342:133–49.
Fox EM, Andrade J, Shupnik MA. (2009) Novel actions of estrogen to promote proliferation: integration of cytoplasmic and nuclear pathways. Steroids. 74:622–7.
Kocher O, Comella N, Tognazzi K, Brown LF. (1998) Identification and partial characterization of PDZK1: a novel protein containing PDZ interaction domains. Lab. Invest. 78:117–25.
Kocher O, et al. (2003) Targeted disruption of the PDZK1 gene in mice causes tissue-specific depletion of the high density lipoprotein receptor scavenger receptor class B type I and altered lipoprotein metabolism. J. Biol. Chem. 278:52820–5.
Kocher O, Krieger M. (2009) Role of the adaptor protein PDZK1 in controlling the HDL receptor SR-BI. Curr. Opin. Lipidol. 20:236–41.
Ghosh MG, Thompson DA, Weigel RJ. (2000) PDZK1 and GREB1 are estrogen-regulated genes expressed in hormone-responsive breast cancer. Cancer Res. 60:6367–75.
Dunbier AK, et al. (2010) Relationship between plasma estradiol levels and estrogen-responsive gene expression in estrogen receptor-positive breast cancer in postmenopausal women. J. Clin. Oncol. 28:1161–7.
Flowers JL, et al. (1986) Use of monoclonal antie-strogen receptor antibody to evaluate estrogen receptor content in fine needle aspiration breast biopsies. Ann. Surg. 203:250–4.
Raff T, van der Giet M, Endemann D, Wiederholt T, Paul M. (1997) Design and testing of beta-actin primers for RT-PCR that do not co-amplify processed pseudogenes. Biotechniques. 23:456–60.
Oumouna-Benachour K, et al. (2007) Poly(ADP-ribose) polymerase inhibition reduces atherosclerotic plaque size and promotes factors of plaque stability in apolipoprotein E-deficient mice: effects on macrophage recruitment, nuclear factor-kappaB nuclear translocation, and foam cell death. Circulation. 115:2442–50.
Zerfaoui M, et al. (2010) Poly(ADP-ribose) polymerase-1 is a determining factor in Crm1-mediated nuclear export and retention of p65 NF-kappa B upon TLR4 stimulation. J. Immunol. 185:1894–902.
Song RX, Zhang Z, Chen Y, Bao Y, Santen RJ. (2007) Estrogen signaling via a linear pathway involving insulin-like growth factor I receptor, matrix metalloproteinases, and epidermal growth factor receptor to activate mitogen-activated protein kinase in MCF-7 breast cancer cells. Endocrinology. 148:4091–101.
Ptak A, Gut P, Blachuta M, Rak A, Gregoraszczuk EL. (2009) Direct inhibition of ERK1/2 phosphorylation as a possible mechanism for the antiproliferative action of 3,4-diOH-PCB3 in the MCF-7 cell line. Toxicol. Lett. 190:187–92.
Barrett T, et al. (2007) NCBI GEO: mining tens of millions of expression profiles: database and tools update. Nucleic Acids Res. 35:D760–5.
Boersma BJ, et al. (2008) A stromal gene signature associated with inflammatory breast cancer. Int. J. Cancer. 122:1324–32.
Saddar S, Mineo C, Shaul PW. (2010) Signaling by the high-affinity HDL receptor scavenger receptor B type I. Arterioscler. Thromb. Vasc. Biol. 30:144–50.
Spears M, Bartlett J. (2009) The potential role of estrogen receptors and the SRC family as targets for the treatment of breast cancer. Expert Opin. Ther. Targets. 13:665–74.
Santen RJ, et al. (2009) Estrogen signals via an extra-nuclear pathway involving IGF-1R and EGFR in tamoxifen-sensitive and -resistant breast cancer cells. Steroids. 74:586–94.
Rajhans R, et al. (2008) Modulation of in situ estrogen synthesis by proline-, glutamic acid-, and leucine-rich protein-1: potential estrogen receptor autocrine signaling loop in breast cancer cells. Mol. Endocrinol. 22:649–64.
Miyoshi Y, Murase K, Saito M, Oh K. (2010) Prediction of hormone sensitivity for breast cancers. Breast Cancer. 17:86–91.
Miller WR, O’Neill J. (1987) The importance of local synthesis of estrogen within the breast. Steroids. 50:537–48.
Frasor J, et al. (2004) Selective estrogen receptor modulators: discrimination of agonistic versus antagonistic activities by gene expression profiling in breast cancer cells. Cancer Res. 64:1522–33.
Mackay A, et al. (2007) Molecular response to aromatase inhibitor treatment in primary breast cancer. Breast Cancer Res. 9:R37.
Englert NA, Spink BC, Spink DC. (2011) Persistent and non-persistent changes in gene expression result from long-term estrogen exposure of MCF-7 breast cancer cells. J. Steroid. Biochem. Mol. Biol. 123:140–50.
Saceda M, et al. (1988) Regulation of the estrogen receptor in MCF-7 cells by estradiol. Mol. Endocrimol. 2:1157–62.
Shupnik MA. (2004) Crosstalk between steroid receptors and the c-Src-receptor tyrosine kinase pathways: implications for cell proliferation. Oncogene. 23:7979–89.
Lefevre G, et al. (2003) Opposite long-term regulation of c-Myc and p27Kip1 through overactivation of Raf-1 and the MEK/ERK module in proliferating human choroidal melanoma cells. Oncogene. 22:8813–22.
Zhu W, et al. (2008) The scavenger receptor class B type I adaptor protein PDZK1 maintains endothelial monolayer integrity. Circ. Res. 102:480–7.
This work was supported in part by grant RSG-116608 from the American Cancer Society and grant HL072889 from the National Institutes of Health (NIH) as well as funds from the Louisiana Cancer Research Consortium (New Orleans, LA, USA) to AH Boulares. This work was also supported in part by grant P20GM103501 (NIH/COBRE) to AS Naura (overall principal investigator: A Ochoa) and by grant PJ008047 from the Next-Generation BioGreen 21 Program Rural Development Administration (Republic of Korea) to JG Kim.
HK and ZYAE contributed equally to this work.
Electronic supplementary material
About this article
Cite this article
Kim, H., Abd Elmageed, Z.Y., Ju, J. et al. PDZK1 Is a Novel Factor in Breast Cancer That Is Indirectly Regulated by Estrogen through IGF-1R and Promotes Estrogen-Mediated Growth. Mol Med 19, 253–262 (2013) doi:10.2119/molmed.2011.00001