|Home | About | Journals | Submit | Contact Us | Français|
To determine the presence, differential expression and regulation of epidermal growth factor-containing fibulin-like extracellular matrix protein 1 (EFEMP1) in uterine leiomyomas.
Laboratory in vivo and in vitro study using human leiomyoma and myometrial tissue and primary cells.
Academic medical center
Leiomyoma and myometrial tissue samples and cultured cells
5-Aza-2′ deoxycytidine (5-Aza-dC) treatment.
Fold change difference between EFEMP1 and fibulin-3 expression in leiomyoma tissue and cells compared to matched myometrial samples, and fold-change difference in EFEMP1 expression with 5-Aza-dC treatment.
In vivo, EFEMP1 expression was 3.19 fold higher in myometrial tissue versus leiomyoma tissue. EFEMP1 expression in vitro was 5.03 fold higher in myometrial cells versus leiomyoma cells. Western blot and IHC staining of tissue and cells confirmed similar findings in protein expression. Treatment of leiomyoma cells with 5-Aza-dC resulted in increased expression of EFEMP1 in vitro.
The EFEMP1 gene and its protein product, fibulin-3, are both significantly downregulated in leiomyoma as compared to myometrium when studied both in vivo and in vitro. The increase in EFEMP1 expression in leiomyoma cells with the 5-Aza-dC treatment suggest that differential methylation is, in part, responsible for the differences seen in gene expression.
Uterine leiomyomas or “fibroids” are benign, sex steroid-sensitive, smooth muscle tumors of the uterus. Leiomyomas occur in as many as 30–50% of reproductive age women and have an overall cumulative incidence of 70% by the age of 50, making them the most common benign tumors in women (1). Symptoms associated with leiomyomas include abnormal uterine bleeding, pelvic pain/pressure, increased abdominal girth and recurrent pregnancy loss, all of which reduce quality of life (2, 3). These symptoms are why leiomyomas remain the leading cause of hysterectomy in the United States, and cost the US health system an estimated $34.4 billion annually (4). Despite their prevalence and public health impact, the etiology of leiomyomas remains unclear.
What we do know is that the extra cellular matrix (ECM) of leiomyomas is much more extensive than adjacent myometrium and is shown to be quantitatively and qualitatively different. While collagens have been well studied in leiomyomas and myometrium, fibulins as a class have not been well characterized. Fibulins are a class of seven secreted extracellular glycoproteins, characterized by tandem calcium-binding epidermal growth factor (EGF)-like domains and a unique C-terminal fibulin structure (5). As a whole, fibulins have been widely studied for their involvement in tumor biology and are specifically implicated for their roles in cell morphology, growth, adhesion and motility (6). Fibulin-3 is encoded by the gene EFEMP1 (epidermal-growth factor-containing fibulin-like extracellular matrix protein 1) and has been suggested to have paradoxical effects on tumor biology. Some evidence suggests that fibulin-3 may act as an antagonist of angiogenesis (7), supporting studies that demonstrate that EFEMP1 downregulation and therefore a loss of fibulin-3 protein may be associated with increased tumor angiogenesis in several cancers (8–11). However, other studies have shown that increased expression of EFEMP1/fibulin-3 may promote tumor growth in pancreatic adenocarcinoma (12) as well as cervical cancer (13).
To our knowledge, neither the presence nor expression pattern of EFEMP1 or fibulin-3 has been formally investigated in leiomyomas. In this paper we will explore the expression and regulation of EFEMP1 and fibulin-3 in leiomyomas and myometrium. We hypothesize that similar to other solid tumors, EFEMP1 is differentially expressed in leiomyomas versus myometrium, both in vivo and in vitro.
Leiomyoma and matched myometrial tissue were collected from subjects (n=20) undergoing hysterectomy for uterine leiomyomata at Northwestern Memorial Hospital. The age ranges of the subjects were 35–52 years old (44.5 ± 4.69 years; mean ± SEM). All subjects were premenopausal non-smokers and none were taking hormonal medication within three months of surgery. Written informed consent was obtained from subjects. The study protocol was approved by the Institutional Review Boards at Northwestern University.
Tissue samples were taken from the operating room to the Pathology Department within 30 minutes of being removed from the patient. Resected leiomyomata ranged in size from 3.5 to 12 cm in diameter and were both subserosal and intramural in nature. The pathologist provided tissue samples as follows: Leiomyoma samples were obtained within 1–2 cm of the outer capsule of the leiomyoma. Myometrial samples were obtained within 2cm of the resected leiomyoma. The tissues were rinsed three times in cold PBS, and transported back to the lab in DMEM/F-12 media on ice. The tissues were then processed for cell culture, paraffin embedding, flash-freezing, protein isolation or nucleic acid isolation as described below.
Leiomyoma and myometrial tissue samples were kept in RNA later, (Invitrogen, Carlsbad, CA) stored at −20° C and, when needed, were homogenized in RLT buffer (Qiagen, Hilden, Germany) using a Polytron PT 2100 homogenizer (Brinkmann Instruments, Westbury, NY). Insoluble material was then removed through centrifugation at 5000rpm for 10 minutes. The supernatant was harvested and total RNA was extracted using the RNeasy mini kit according to the manufacturer’s instructions (Qiagen). RNA concentration and purity were later confirmed by spectrophotometry using NanoDrop ND-1000 (NanoDrop Technologies, Wilmington, DE). RNA integrity was confirmed by gel electrophoresis.
Real-time polymerase chain reaction (RT-PCR) was used to amplify the mRNA expression of EFEMP1 through the ABI Taqman gene expression system and the ABI7900 sequence detection system (Applied Biosystems, Foster City, CA). The mRNA expression levels are reported as fold change, as calculated using the 2−ΔΔCT methodology (14).
The procedure for isolation of primary cells from leiomyoma and myometrium has been previously described (15). Briefly, fresh leiomyoma and myometrial tissue was rinsed in PBS, cut into 2–3 mm cubed pieces and incubated with collagenase and DNase for 8 to 10 hours in a warm shaking incubator. The digested lysate was then strained to remove any undigested tissue and debris. The lysate containing the cells was then centrifuged for 10 minutes at 5000 rpms at room temperature. The resulting cell pellet was rinsed three times with DMEM/F-12 media containing 10% FBS (Gibco, Grand Island, NY) and 1% antibiotic-antimycotic mixture. The cells were counted and plated at the desired concentration in the same media. Once the plates were confluent, subcultures of each type of tissue were trypsinized and passaged up to two times with 0.1% trypsin-.5% EDTA.
Primary cells were treated with 5-Aza-dC, an epigenetic modifier that inhibits DNA methyltransferase activity. Primary leiomyoma cells were placed in serum-free DMEM/F-12 overnight and then treated with varying concentrations of 5-Aza-dC (0, 1, 5, 10, 20 μM) for five days. All 5-Aza-dC containing media were changed on a daily basis. After five days, RNA and protein were isolated.
For protein extraction from tissue, flash-frozen leiomyoma and myometrial tissues were removed from the −80° C freezer, ground in liquid nitrogen, and lysed in T-PER® protein extraction reagent (Thermo Fischer Scientific, Franklin, MA). For protein isolation from primary cultured cell, primary cells were washed in ice-cold PBS and suspended in T-PER® protein extraction reagent. The protein was extracted following the manufacturer’s protocol and the concentration was quantified using the Bicinchoninic (BCA) assay protein (Pierce Biotechnology, Rockford, IL) as indicated by the manufacturer’s instruction. Approximately 25μg of protein was diluted with reducing 4X lithium dodecyl sulfate sample buffer (Life Technologies, Grand Island, NY), electrophoresed on 4–12 Tris-HCL gels (Life Technologies), and transferred onto nitrocellulose membranes. The membranes were blocked with 5% milk in Tris-buffered saline and 0.2% Tween 20 and incubated with EFEMP1 antibody overnight at 4°C. (Santa Cruz Biotechnology, Santa Cruz, CA). An anti β-actin antibody was used as a loading control (Sigma-Aldrich, St. Louis, MO). An enhanced chemiluminescence system, ECL Western Blotting (Amersham Pharmacia Biotech, Piscataway, NJ), was used to detect proteins on gel substrates.
Antibodies used for this study included Fibulin-3 antibodies (Santa Cruz Biotechnology). Tissues were sectioned 4 μm in thickness. After deparaffinization and antigen retrieval, all immunohistochemical staining was performed on a Ventana Nexus automated system (Tucson, Arizona). In brief, endogenous peroxidase activity was blocked with 3% hydrogen peroxide. Primary antibodies were detected using standard biotinylated anti-mouse or anti-rabbit secondary antibodies.
Leiomyoma and myometrium cells were grown on glass chamber slides (BD Falcon, Pittston, PA). Cells were fixed for one hour with 4% paraformaldehyde (Electron Microscopy Sciences, Hatfield, PA). Cells were then permeabolized for 10 minutes with 0.2% Triton-X (Sigma-Aldrich). Cells were then blocked with a 10% FBS solution for 1 hour. Subsequently, the Fibulin-3 primary antibody made in a 1% FBS solution at a concentration of 1:50 was added to each sample chamber and was placed overnight in a 4° room on a shaker. A fluorescein secondary Alexa Fluor 488 goat anti-mouse IgG (Life Technologies) was then used in each chamber at a concentration of 1:200 and the cells were allowed to shake for two hours at room temp. A 1:200 solution of monoclonal Anti-Actin, α-smooth muscle-Cy3 (Sigma-Aldrich) in 1% FBS was used on the cells and they were allowed to shake at room temp for two hours. After the smooth muscle actin, a DAPI (Sigma-Aldrich) solution made with 1% FBS was placed in each chamber and was allowed to sit for five minutes. Cells were then mounted with Permafluor mountant (Thermo Fisher Scientific) and visualized using an automated upright microscope system, Leica DM5000 B (Leica microsystems, Buffalo Grove, IL).
An analysis of relative fold change of EFEMP1 and fibulin-3 expression was conducted for both in vivo and in vitro procedures. Student t-tests compared myometrium to matched leiomyoma tissue and cell cultures. Analysis of variance was used to compare 5-Aza-dC treatment responses. P-values <0.05 were considered significant. Relevant data are presented as mean ± standard error of the mean (SEM) and each experiment was performed at least three times.
Real time PCR results definitively demonstrate that EFEMP1 gene expression is significantly downregulated in leiomyoma versus matched myometrium in vivo (0.31±0.07, p<0.0001; Figure 1). Immunohistochemical staining of matched leiomyoma and myometrial tissue pairs further demonstrate that the expression of the protein product of EFEMP1, fibulin-3, is also significantly downregulated in vivo in leiomyoma versus myometrium (Figure 1). Results also demonstrate both cytoplasmic and nuclear staining.
Because genetic and protein findings seen in vivo are often lost in transition to in vitro models, we sought to assess EFEMP1 and fibulin-3 in vitro as well. Real-time PCR demonstrates a markedly decreased expression of EFEMP1 in leiomyoma cells versus matched myometrial cells (0.20 ± 0.03, p=0.002; Figure 2). Western blot staining of matched staining cells from three subjects also confirms that fibulin-3 protein is markedly downregulated in leiomyoma cells versus myometrial cells. Additional immunofluorescence staining of matched myometrial and leiomyoma cells reveal findings consistent with the western blot results. Leiomyoma cells have markedly decreased staining for fibulin-3 (Figure 3).
After five days of treatment with 5-Aza-dC, EFEMP1 expression increased significantly relative to vehicle at the lowest dose tested. There were no significant differences between the treatment responses themselves (Figure 4).
In this study, we used in vivo and in vitro methods to determine that both EFEMP1 gene expression and fibulin-3 protein levels are significantly downregulated in leiomyoma versus myometrium, and that differential methylation may be responsible for the observed downregulation. These findings were confirmed using RT-PCR, western blot, immunohistochemistry and immunofluorescence. 5-Aza-dC treatments were conducted to determine whether methylation status played a role in gene expression. These experiments resulted in an increase of EFEMP1 expression in vitro. To our knowledge, this study is the first to report on the expression of EFEMP1 and its regulation in benign solid tumors in general, and in leiomyoma specifically.
Though no other studies have looked at the relationship between EFEMP1 and leiomyoma growth, our findings are consistent with observations of EFEMP1 downregulation in a variety of malignancies including prostate, hepatocellular and sporadic breast cancers (8–11, 16, 17). Both Kim et al. and Nomoto et al. used microarray methods to identify EFEMP1 as a potential DNA biomarker for prostate cancer and hepatocellular cancer, respectively (8, 16). Sadr-Nabavi et al. determined a similar outcome in sporadic breast cancer using RNA microarray expression analyses (10). Of note, our findings also correlate with demonstrations of EFEMP1 downregulation reported in instances of endometrial cancer by Yang et al. (18). According to this study, the decrease in EFEMP1 expression is mainly regulated by promoter hypermethylation, a finding consistent with our preliminary results that suggest differential methylation patterns in leiomyomata. In addition, a similar treatment of endometrial cancer cells with 5-aza-dC also led to a reversal of promoter methylation status in that experiment.
While consistent with many studies, our findings were discordant with observations of EFEMP1 expression in pancreatic adenocarcinoma (12) and cervical cancer (13), in which increased EFEMP1 expression was shown to promote tumor angiogenesis. This shows that there is high variability regarding the role of EFEMP1 in tumor development. Some cancer types also have been met with mixed results concerning the involvement of EFEMP1, particularly in the case of ovarian cancer cell lines. Press et al. determined that EFEMP1 expression in BRCA1-mutated ovarian carcinoma in tubal epithelium was downregulated (11). Conversely, findings presented by Januchowski et al. demonstrated significant EFEMP1 upregulation in A2780 ovarian cancer cell lines (19). It is speculated that the EFEMP1 upregulation seen in this experiment may have influenced anti-apoptotic properties of ovarian tumor cells, though the pathway for this mechanism is unclear. These inconsistencies in EFEMP1 regulation could be due, in part, to the cellular differences associated with various cancer types, and differences in downstream cellular receptor activation and signaling (20). The environments in which experimental procedures were carried out could also have had a potential impact on results.
In concert with EFEMP1 gene downregulation, our findings were also consistent with previous studies on fibulin-3 expression in solid tumors in which fibulin-3 levels were found to be lower in cancerous cells compared to their normal counterparts (9, 21). Though our experiments only noted fold-change of fibulin-3 levels, other studies have shown that when present, fibulin-3 may serve as a protective factor against tumor growth by targeting metalloproteinase (MMPs) pathways involved in cancer pathogenesis, particularly in lung cancer (22–24). Such mechanisms were also evident in noncancerous conditions including osteoarthritis and macular degenerative diseases (25, 26). There is also evidence that suggests that EFEMP1 and fibulin-3 play a role in the inhibition of signaling pathways involving epidermal growth factor receptor (EGFR), protein kinase B (AKT) and transforming growth factor-beta (TGF-β) (20, 27). In fact, aberrant activity of both EGF and AKT pathways have previously been implicated in leiomyoma (28, 29). Future studies may benefit from exploring similar mechanism interactions between EFEMP1, fibulins and these distinct pathways in leiomyoma tissue.
Promoter hypermethylation and associated gene silencing have been demonstrated in other leiomyoma tumor suppressor genes as well. Specifically, Navarro et al. used genome-wide analyses of DNA-methylation and mRNA expression in leiomyoma tissue to show that genes KLF11, DLEC1, and KRT19 were significantly downregulated in leiomyoma versus normal uterine smooth muscle tissue (30). Similar findings were demonstrated by Yang et al., who further reported on possible involvement of both histone deacetylase and histone methyltransferase in leiomyoma pathogenesis, as well as the existence of side population (SP) cells with “tumor-initiating properties” (31). Our findings demonstrating an increase in EFEMP1 expression with AZA treatment could lead to further identification of research targets and potential diagnostic biomarkers for leiomyoma and warrants additional research in order to appreciate the complexity of leiomyoma biology.
The strengths of this study include its use of human tissue, its consistency of extraction, isolation and treatment protocols among all tissue specimens and primary cell cultures and its large sample size. Limitations of the study include use of primary cells instead of cell lines. This study also was not powered to account for differential gene expression in subsets of fibroids based on location or size. Additionally, though we achieved consistent and reproducible results with the early passage procedures of primary leiomyoma cells, we acknowledge that the monoclonal nature of these tumors could allow for potentially significant subject-to-subject variation. We attempted to control for this by using uniform tissue extraction protocols developed by our institution (32, 33).
In conclusion, EFEMP1 and its protein product fibulin-3 are downregulated in leiomyoma versus myometrial tissue in vivo and in vitro. These findings are consistent with other studies conducted on solid malignant tumors and suggest a common pathological mechanism. Further studies will need to assess the role of EFEMP1 in leiomyoma tumorigenesis. In particular, EFEMP1 and its relation to signal inhibition of various molecular pathways, as well as the mechanism by which it is differentially expressed in leiomyoma versus myometrium can be potentially beneficial in developing long-term, non-hormonal therapeutic and preventative options for this burdensome disease.
Support for this study was provided by NIH K12HD050121 Women’s Reproductive Health Research Scholar Program at Northwestern (EEM/SEB), Feinberg School of Medicine – Northwestern University, Northwestern Memorial Hospital, Robert Wood Johnson Foundation (EEM).
Disclosures: Dr. Marsh has served on an advisory board for Abbvie. These data have been presented in part as an oral presentation at the 59th Annual Meeting of The Society for Gynecologic Investigation.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.