Differential Expression of Receptor Tyrosine Kinases (RTKs) and IGF-I Pathway Activation in Human Uterine Leiomyomas

Abstract

Uterine leiomyomas (fibroids) are benign tumors that are prevalent in women of reproductive age. Research suggests that activated receptor tyrosine kinases (RTKs) play an important role in the enhanced proliferation observed in fibroids. In this study, a phospho-RTK array technique was used to detect RTK activity in leiomyomas compared with myometrial tissue. We found that fifteen out of seventeen RTKs evaluated in this study were highly expressed (P < 0.02–0.03) in the leiomyomas, and included the IGF-I/IGF-IR, EGF/EGFR, FGF/FGF-R, HGF/HGF-R, and PDGF/PDGF-R gene families. Due to the higher protein levels of IGF-IR observed in leiomyomas by us in earlier studies, we decided to focus on the activation of the IGF-IR, its downstream effectors, and MAPKp44/42 to confirm our earlier findings; and validate the significance of the increased IGF-IR phosphorylation observed by RTK array analysis in this study. We used immunolocalization, western blot, or immunoprecipitation studies and confirmed that leiomyomas overexpressed IGF-IRβ and phosphorylated IGF-IRβ. Additionally, we showed that the downstream effectors, Shc, Grb2, and MAPKp44/42 (P < 0.02–0.001) were also overexpressed and involved in IGF-IR signaling in these tumors, while IRS-I, PI3K, and AKT were not. In vitro studies showed that IGF-I (100 ng/mL) increased the proliferation of uterine leiomyoma cells (UtLM) (P < 0.0001), and that phosphorylated IGF-IRβ, Shc, and MAPKp44/42 were also overexpressed in IGF-I-treated UtLM cells (P < 0.05), similar to the tissue findings. A neutralizing antibody against the IGF-IRβ blocked these effects. These data indicate that overexpression of RTKs and, in particular, activation of the IGF-IR signaling pathway through Shc/Grb2/MAPK are important in mediating uterine leiomyoma growth. These data may provide new anti-tumor targets for noninvasive treatment of fibroids.

Introduction

Uterine leiomyomas (fibroids) are benign neoplasms of the myometrium that are prevalent in reproductive-aged women in the United States (1). Some fibroids are asymptomatic; however, many can cause pelvic pain, menstrual bleeding, and infertility. In the United States, fibroids represent a tremendous public health burden for women and a great economic cost to society (2). Treatment options for leiomyomas are currently very limited; surgery remains the main form of treatment (3).

Receptor tyrosine kinases (RTKs) are the main mediators of the signaling network that transmit extracellular signals into the cell, and control cellular differentiation and proliferation (4). Recent and rapid advances in the understanding of cellular signaling by RTKs in normal and tumor cells have brought to light the potential of RTKs as selective anti-tumor targets. Their activity is normally tightly controlled and regulated; however, overexpression of RTK proteins or abnormal stimulation by autocrine growth factor loops contribute to constitutive RTK signaling, resulting in dysregulated cell growth and tumor formation (4). Uterine leiomyomas express many types of growth factors (5,6). Those factors may foster leiomyoma growth through local paracrine and/or autocrine mechanisms (5,7). However, the association between upregulation of various growth factor RTKs and leiomyoma development is not fully understood. In this study, we obtained uterine leiomyomas and patient-matched myometrial tissue from premenopausal women to investigate the differential expression of growth factor RTKs involved in cell mitogenesis by using a RTK array technique.

Previous studies in our laboratory (5,8) and by others (7,9–11) have indicated that the IGF-I pathway plays an important role in uterine leiomyoma development and growth. IGF-I is a single-chain polypeptide whose structure is highly similar to that of proinsulin. It can bind either to its specific cell surface receptor IGF-IR or to the closely related insulin receptor (IR), although it has a much higher affinity for its own receptor (12). IGF-IR is one of the major receptor tyrosine kinase proteins, and its central role in the IGF-I family in the regulation of both physiological and pathological growth processes has been firmly established (13). The IGF-IR utilizes IRS-I/Shc as immediate downstream adaptors which can ultimately lead to the activation of the IRS/PI3K/AKT cell survival pathway and/or the Shc/Ras/Grb2/MAP kinase cell proliferation pathway (14). Based on our previous studies, and to confirm the overexpression level of phosphorylated IGF-IR observed in the RTK arrays in this study, we chose to evaluate the IGF-IR and its pathway activation in leiomyoma and myometrial tissues further, and to assess IGF-IR signaling in uterine leiomyoma cells in culture.

Materials and Methods

Subjects

Uterine leiomyoma and patient-matched myometrial tissue samples were collected from eight to ten women ranging from 41 to 49 years of age who underwent hysterectomy for symptomatic leiomyomas. All subjects had taken no hormonal medication within at least 3 months prior to hysterectomy. Informed consent was obtained, and the Institutional Review Board (IRB) of the NIEHS, NIH approved the study. All leiomyomas and unaffected myometrial samples were confirmed by histological evaluation. The endometrium from each uterus was evaluated to determine the menstrual cycle phase, with all tumors taken from women in the proliferative phase of the menstrual cycle.

Phosphorylation of Receptor Tyrosine Kinases (RTKs) Array

Expression of phosphorylated growth factor receptor tyrosine kinases (RTKs) was detected using the Proteome Profiler Array Kit (R&D Systems, Minneapolis, MN, USA). Samples of 50 to 100 mg of frozen leiomyoma and myometrial tissue from each of ten patients were collected in cold solubilization buffer (1% Triton X-100, 1 mM sodium vanadate, 1 mM sodium fluoride, 0.05 mM sodium molyb-date, 20 µg/mL aprotinin, 20 µg/mL leupeptin, 4 µg/mL (4-amidinophenyl) methane sulfonyl fluoride, 150 mM sodium chloride in 50 mM Tris-HCl, pH 7.4). The samples were minced and then homogenized on ice using a 30 s burst of a homogenizer at its highest setting, followed by centrifugation at 16,000g for 5 min at 4°C. The supernatants were collected and stored at −80°C. Three-hundred µg of pooled total protein from ten leiomyomas and patient-matched myometrial tissue (30 µg from each patient sample) were incubated with RTK array membranes spotted with various anti-phospho-RTK antibodies. The procedures were performed according to the manufacturer’s protocol.

Histology and Immunohistochemistry

The tissues from eight of the ten women from the RTK studies were fixed overnight in 10% neutral buffered formalin, embedded in paraffin, sectioned at 6 µm and mounted onto charged glass slides (ProbeOn Plus, Fisher Scientific, Pittsburgh, PA, USA) for immunohistochemical staining. Tissues were deparaffinized with xylene and rehydrated with ethanol. Endogenous peroxidase activity was blocked with 0.3% H₂O₂ (0.1 mL 30% hydrogen peroxide [Fisher Scientific, Fair Lawn, NJ, USA]) for 30 min. Tissues were pressurized and depressurized by a decloaker and blocked for 30 min with normal serum (Vector Laboratories, Burlingame, CA, USA). Tissues were incubated overnight at 4°C with respective diluted antibodies: 1:25 for IGF-I and 1:75 for IGF-IRβ (Santa Cruz Biotechnology, Santa Cruz, CA, USA), 1:50 for IR (Upstate Biotechnology, Lake Placid, NY, USA), and 1:50 for phospho-MAPKp44/42 and phospho-AKT (Cell Signaling Technology, Beverly, MA, USA). The same concentrations of nonimmune rabbit or goat serum (Santa Cruz Biotechnology) were used as negative controls. Secondary antibodies were applied for 60 min (one drop of biotinylated anti-rabbit or goat IgG, respectively, with three drops normal goat or rabbit serum, respectively). Tissues were labeled for 60 min with two drops of Avidin + two drops biotinylated Enzyme (Vector Laboratories). DAKO Liquid DAB (3,3′-diaminobenzidine tetrahydrachloride) large volume substrate-chromogen system (DAKO Corporation, Carpinteria, CA, USA) was applied for 6 min in the dark. Tissues were counterstained for 45 s with Mayer’s hematoxylin (Polyscientific, Bay Shore, NY, USA), dehydrated with ethanol and xylene and mounted with Permount, then coverslipped.

Cell Culture Studies and MTT Assay

Human uterine leiomyoma (UtLM) cells (GM10964) were purchased from Coriell Institute for Medical Research (Camden, NJ, USA), and grown in UtLM media containing Minimum Essential Medium (MEM; Gibco Life Technologies, Grand Island, NY, USA), 1X vitamins, 1X non-essential amino acids, 1X essential amino acids, 2X L-glutamine (Gibco Life Technologies) and 20% Fetal Bovine Serum (FBS; Sigma, St. Louis, MO, USA). UtLM cells were seeded into 96-well plates at 5,000 cells/well in growth medium. The media were switched to phenol red-free Dulbecco’s Modified Eagel’s Medium (DMEM) with 10% FBS, or 10% charcoal/dextran-treated FBS after 24 h, and switched to phenol red-free DMEM only after 48 h. The cells were treated 24 h later with media containing serum (FBS) + DMEM, or serum free DMEM with or without LongR-IGF-I (IGF-I; GroPep Limited, The Barton SA, Australia), or charcoal/dextran treated serum + DMEM with or without IGF-I peptide. The IGF-I treatment media consisted of 100 ng/mL IGF-I and 0.1% bovine serum albumin. The cells were treated with IGF-I every 3 days and were counted on each of day 0, 3, 9, 12, and 15 using a commercially available MTT assay kit (Cell Titer 96 Aqueous One Solution Cell Proliferation Assay, Promega, Madison, WI, USA) according to the manufacturer’s protocol. Plates were consistently read after 1.5 h of incubation with the assay reagent. The study was repeated at least three times; statistical analysis was used to test for differences between IGF-I treatments and their respective non-treated controls between serum (FBS) and serum free medium conditions, and between IGF-I treatments before and after anti-IGF-IR blocking. Readings from the wells with no cells were subtracted as background.

To assess early activation of a specific pathway, in particular the IGF-IR pathway, cells must be tested in a serumfree environment to eliminate the possibility of activation by cytokines in the serum. UtLM cells were grown in UtLM media until 70% confluent, and maintained in charcoal/dextran treated FBS for 3 days and serum free for 1 day. The cells were treated with IGF-I peptide (100 ng/mL) for 0, 5, 10, 30, and 60 min separately and cell lysates were harvested and stored at −80°C prior to western blotting. For IGF-IR blocking, the cells were blocked with anti-goat-hIGF-IR (2 µg/mL) (R&D Systems) for 2h, then cells were induced with LongR-IGF-I peptide (100 ng/mL) for 0, 5, 10, 30 and 60 min. The cell lysates were harvested and stored at −80°C prior to western blotting.

Western Blotting of Tissue Homogenates and Cells

Atotal of 50–100 mg of frozen leiomyoma and myometrial tissues from each of the same eight patients as those in RTKs and Immunohistochemistry studies were collected in cold solubilization buffer (the same as described in the RTKs array procedure). The supernatants containing protein lysate were collected and stored at −80°C for western blot analysis.

For the in vitro studies, UtLM cells were harvested in RIPA buffer (phosphate-buffered saline, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing protease inhibitors aprotinin (10 µg/mL), leupeptin (10 µg/mL) and phenymethanesulfonyl fluoride (1 mM), after treatment with IGF-I at different time points. The cell lysates were centrifuged and stored at −80°C prior to Western blot analysis as previously described (15). The western blots were incubated overnight with a specific diluted primary antibody: 1:1000 for total and phosphorylated IGF-IRβ, IRβ, MAPKp44/42, IRS-I, Shc and AKT, and PI3K (Cell Signaling Technology), and 1:300 for Grb2 (Santa Cruz Biotechnology). A densitometer (Fluor ChemTM8900, α Innotech, San Leandro, CA, USA) was used for quantitation of band intensities.

Immunoprecipitation

Atotal of 300 µg of total protein from tissue homogenates (as described in the RTKs array procedure) in 500 µl solubilization buffer was precleared by incubation, with 50 µL of Protein A-sepharose CL-4B Beads (Amersham Pharmacia Biotech, Piscataway, NJ, USA) at 4°C for 30 min. Precleared samples were incubated for 2 h with rabbit polyclonal IGF-IRβ or phosphorylated Shc at 4°C. The antigen-antibody complex was captured by incubation with 50 µl of protein A-sepharose CL 4B overnight at 4°C. The beads were washed three to four times in solubilization buffer. Immune complexes were eluted from the beads using 2X laemmli sample buffer (BIO-RAD, Hercules CA, USA) with 5% 2-mercaptoethanol added to the buffer before use.

A Seize Primary Immunoprecipitation Kit (Pierce Biotechnology, Rockford, IL, USA) was used to keep the antigen free from antibody contamination. The kit was applied because IgG of anti-IGF-IRβ, which was used to pull down Shc, has the same molecular weight as the target protein Shc and this kit allows the immunoglobulin to remain adherent to aminolink Plus Gel following elution. Amounts of 200 µg of antibody IGF-IRβ were coupled to 400 µl of 50% of aminolink Plus Gel Slurry in coupling buffer overnight at 4°C. The coupled gel and antibody complex were divided into eight tubes, and each tube incubated with 300 µg of the total protein in binding buffer overnight at 4°C. The gel was washed three times with washing buffer (manufacturer’s protocol). Only the antigen in the antigen-antibody complexes was eluted by the elution buffer (all buffers supplied in the Kit).

Statistical Analysis

Mann-Whitney tests (16) were used to determine statistically significant differences between leiomyoma and myometrial tissue for each RTK receptor dot intensity value. Mann-Whitney tests were also used to compare the western blot intensity value of leiomyoma and myometrial tissue with respect to total and phospho-IGF-IRβ, IRS-I, Shc, AKT, and MAPKp44/42, and their ratios; and to compare myometrial tissue with leiomyoma tissue for intensity values of PI3K, IR, and immunoprecipitated values of Shc, IRS-I, and Grb2. For in vitro studies, the Wilcoxon signed rank test was used to test for differences between IGF-I treatments and their respective non-treatment controls, between serum (FBS) and serum-free medium conditions, and between with and without anti-IGF-IR blocking with IGF-I treatments. Also, Mann-Whitney tests were used to compare the western intensity values of the IGF-I treatment and non-treatment groups, and the IGF-IR blocked and non-blocked groups. The analysis of RTKs was based on four replicates of data. The western blot densitometry procedure from tissue was based on data from eight patients. For the in vitro IGF-I treatment MTT assay, 8 or 16 wells in 96 wells plates were treated as replicates for each treatment condition, and the assay was repeated at least three times. The western blotting analysis of IGF-I treated UtLM cells was based on three replicates of data.

Results

Differential Expression of Growth Factor RTKs in Leiomyoma and Myometrial Tissue

We performed phospho-receptor tyrosine kinase arrays to determine the expression profile of growth factor RTKs in leiomyoma and patient-matched myometrial tissues. The degree of differential expression of growth factor RTK proteins between leiomyoma and myometrial tissues was quantitated by dot blot (17 phosphorylated growth factor RTK antibodies evaluated on the membrane). Ten leiomyomas and patient-matched myometrial samples were collected, and pooled tumor and myometrial protein samples were incubated on separate RTK membranes followed by densitometric measurements of dot blots.

The various growth factor receptor tyrosine kinases were differentially expressed in leiomyoma and myometrial tissues (Figure 1A). Of 17 RTKs evaluated on the array membrane for this study, significantly greater expression levels were observed with 15 of the phosphorylated RTKs in the leiomyomas compared with myometrial samples (P ≤ 0.02–0.03) (Figure 1B). The significant mean (n = 4 dots) fold changes of expression of RTKs in leiomyoma tissue over that of myometrial tissue were: EGFR (1.9), ErbB4 (3.2), FGFR1 (3.0), FGFR2α (3.3), FGFR3 (2.9) FGFR4 (3.8), InsulinR (2.4), IGF-IR (2.7), HGF-R (3.5), MSP-R (2.7), PDGF-Rα (3.8), PDGF-Rβ (4.8), SCF-R (7.9), Fit3 (6.2), and M-CSF-R (6.5). These receptors basically belong to the EGF, Insulin, IGF-I, FGF, HGF, and PDGF growth factor gene families—all of which are involved in cell proliferation and differentiation.

Figure 1

Differential expression of phosphorylated growth factor RTKs in leiomyoma and myometrial tissues.

(A, B) Out of 17 RTKs analyzed in the Phosphorylation of Receptor Tyrosine Kinases Array, 15 growth factor RTKs were highly expressed (P < 0.02–0.03) in leiomyoma (L) tissue compared with myometrial (M) tissue. These RTKs belong to the EGF, FGF, IGF-I, HGF, and PDGF growth factor receptor gene families. The array was done on pooled samples of ten leiomyoma tissue lysates and ten patient-matched myometrial tissue lysates from ten patients, and the dot and triangle symbols represent the dot blot (n = 4) intensity values for pooled myometrial and leiomyoma samples, respectively.

Immunohistochemical Localization of IGF-I, IGF-IRβ, IRβ, and MAP Kinase (MAPK)p44/42 in Leiomyoma and Myometrial Tissue

Based on our previous studies, and to confirm the overexpression level of phosphorylated IGF-IR observed in the RTKs array in this study, we chose to further evaluate the IGF-IR and its pathway activation in leiomyoma and myometrial tissue. Immunohistochemical localization of IGF-I, IGF-IRβ, phospho-MAPKp44/42, and phospho-AKT in leiomyoma and myometrial tissue was assessed. To check the specificity of the IGF-Rβ antibody, IRβ was also assessed (data not shown).

Immunoexpression of the IGF-I peptide was much more pronounced in leiomyomas than in myometrial tissue. Positive staining for IGF-I peptide in the myometrium was mostly perivascular and minimal (Figure 2A). In the leiomyomas, the cytoplasm of smooth muscle tumor cells and the fibroblasts in the extracellular matrix of the tumor tissue stained moderately to intensely positive for IGF-I; whereas, the smooth muscle cells and interstitial connective tissue of myometrium did not (Figures 2A and 2B). Both tumor and myometrial samples stained positively for IGF-IRβ. However, staining was more intense in the tumor samples and was found mostly in the cytoplasm and cytoplasmic membranes of both the smooth muscle cells in the myometrium and the leiomyoma cells (Figure 2C and 2D). Phospho-MAPKp44/42 staining was nuclear, and was minimal in the myometrium (Figure 2E). However, the nuclei of fibroblasts in the extracellular matrix regions and smooth muscle tumor cells in the leiomyoma tissue showed intense positive nuclear staining for phospho-MAPKp44/42 (Figure 2F). Interestingly, there was no difference in staining between leiomyoma and myometrial tissues for phospho-AKT in most of the samples; however, two out of eight (25%) of the tumor samples showed slightly more intense intranuclear and cytoplasmic phospho-AKT staining compared with patient-matched myometrial tissue (Figures 2G and 2H). Using immunohistochemistry, both leiomyoma and myometrial samples were only minimally positive for the insulin receptor (data not shown).

Figure 2

Immunolocalization of IGF-I, IGF-IRβ, phospho-MAP kinase, and phospho-AKT in leiomyoma and myometrial tissues.

The immunohistochemical staining was performed on leiomyoma and matched myometrial tissue samples from eight patients with comparable results. Both myometrial (M) and leiomyoma (L) tissues stained positive for IGF-I, IGF-IRβ, MAPK, and AKT; however, IGF-I (A, B) was localized to the cytoplasm of smooth muscle cells and within the fibroblasts in the extracellular matrix in the tumors, whereas, positive staining in myometrial tissue was mostly perivascular and minimal. (C, D) IGF-IRβ staining was observed in the cytoplasm and cytoplasmic membranes and was more intense in tumor smooth muscle cells. (E, F) Phospho-MAPKp44/42 was strongly positive in nuclei of smooth muscle cells in tumor, but minimal in myometrial tissue. (G, H) There were some staining for phospho-AKT, but no staining difference between myometrial and tumor tissues was observed.

Western Blot Analysis of Phosphorylated IGF-IR and Associated RTK Pathway Proteins in Leiomyoma and Myometrial Tissues

To further investigate the activation of the IGF-IR pathway and to delineate which specific proteins are phosphorylated in leiomyoma tissue compared with myometrium, we performed western blot analysis to assess the status of IGF-IR, IRS-I, Shc, PI3K, Grb2, AKT, and MAPKp44/42 expression in eight leiomyomas and patient-matched myometrial tissue samples. Both total and phosphorylated IGF-IRβ, Shc, and MAPKp44/42 were identified by western blotting in leiomyoma and myometrial tissues (Figure 3A). However, greater phosphorylation of IGF-IRβ, Shc, and MAPKp44/42 was observed in tumor samples, as indicated by the increased intensity of the bands in tumors compared with the bands from myometrial tissue (Figure 3A). Furthermore, the ratios of the mean intensities of bands of phosphorylated over total protein for IGF-IRβ (P < 0.04), Shc (P < 0.05), and MAPKp44/42 (P < 0.02) were significantly higher in tumors compared with patient-matched myometrial samples (Figure 3B). Lower levels of expression of PI3K were observed in both myometrial and leiomyoma tissues (Figure 3A). Although there was expression of IRS-I and AKT in both tissue types (Figure 3A), the ratios of the mean intensities of phosphorylated IRS-I and AKT over total protein levels were not significantly different between leiomyoma and myometrial tissues (Figure 3B).

Figure 3

Western blot analysis of IGFI/IGF-IR pathway activation in leiomyoma and myometrial tissues.

(A,B) Phosphorylated IGF-IRβ (P < 0.04), Shc (P < 0.05), and MAPKp44/42 (P < 0.02) expression levels were significantly higher in leiomyomas (L) compared with myometrial (M) tissue samples. There was some expression of phosphorylated and total IRS-I and AKT, but no significant differences were found between tumor and myometrial tissue samples. There was very minimal expression of PI3K. Western blot analysis was done at least three times in three independent leiomyoma and matched myometrial tissue preparations from eight patients with comparable results. The band images shown represent four of eight patient samples. The phospho/total ratio data was expressed as mean ± SE in eight patient samples.

Immunoprecipitation of Phospho-Shc, Phospho-IRS-1 and Grb2

Immunoprecipitation studies using leiomyoma and patient-matched myometrial tissue lysates were performed to determine which proteins downstream of IGF-IR were activated and associated with the IGF-IR. We found that more phospho-Shc was precipitated to the IGF-IRβ in leiomyoma samples compared with myometrial samples (P < 0.02); however, there was no difference in IRS-I protein associated with the IGF-IRβ in tumor versus myometrial tissue (Figure 4A and 4B). Grb2, a protein immediately downstream of Shc in the MAPK signaling pathway, showed increased precipitation to Shc (P < 0.001) in the leiomyoma samples compared with myometrial tissues (Figure 4A,4B). Our data indicate that phospho-IGF-IRβ and the downstream adaptor and effector proteins phospho-Shc, Grb2, and the MAP kinase pathway were activated in uterine leiomyomas compared with myometrial samples, and more so than the PI3K/AKT signaling pathways.

Figure 4

Immunoprecipitation and western blot analysis of IGF-IR pathway protein expression in leiomyoma and myometrial tissues.

(A) Immunoprecipitation and immunoblots for phospho-Shc or IRS-I, and Grb2 were done to look for an association of phospho-Shc or IRS-I to the IGF-IRβ, and an association of Grb2 to phospho-Shc. (B) Concentrations of phosphorylated Shc associated with IGF-IRβ were significantly higher (P < 0.02) in leiomyoma (L) tissue samples compared with myometrial (M) tissue samples. Levels of Grb2 associated with Shc were also significantly higher (P < 0.001) in leiomyomas compared with myometrial samples. However, there was no difference in levels of phosphorylated IRS-I associated with IGF-IRβ in myometrial and leiomyoma samples. The procedure was performed at least three times in three independent leiomyoma and matched myometrial tissue preparations from eight patients with comparable results. The band intensity of the blots was expressed as means ± SE in eight patient samples.

Functional Effects of IGF-I on UtLM Cell Proliferation

To examine the functional properties of IGF-I on proliferation of uterine leiomyoma (UtLM) cells in vitro, UtLM cells were cultured in DMEM containing IGF-I (100 ng/mL) in the presence or absence of 10% charcoal/dextran-treated (stripped serum) FBS, and in the presence or absence of 10% normal FBS. There were significant differences in proliferation under the different culture conditions as evidenced by using a MTT assay (Figure 5). The cell numbers were significantly higher in the presence of 10% FBS, or presence of IGF-I (100 ng/mL) with 10% charcoal/dextrantreated FBS (stripped serum, SS) compared with cells receiving no FBS (serum free, SF) and no IGF-I at days 3, 9, 12, and 15 (P < 0.0002, P < 0.0001). Surprisingly, UtLM cell growth in SS plus IGF-I medium equaled the growth of UtLM cells cultured in complete FBS medium at 9 days after treatment. In serum-free medium, there was insufficient IGF-I and other growth factors and cytokines needed to maintain UtLM cell survival and promote cell growth. The proliferative effect of IGF-I on UtLM cells was inhibited by a neutralizing antibody against IGF-IRβ (P < 0.0002, Figure 5).

Figure 5

Proliferative effects of IGF-I on uterine leiomyoma (UtLM) cells.

IGF-I (100 ng/mL) increased UtLM cell growth in a time-dependent manner. The UtLM cells had significantly higher growth rates when cultured in media with FBS (P < 0.0001) or with charcoal/dextran stripped (SS) media plus IGF-I (P < 0.0001) compared with UtLM cells cultured in media without FBS (SF), or with charcoal/dextran stripped media without IGF-I, or with IGF-I in the absence of SS. Note that UtLM cell growth in SS plus IGF-I medium equaled UtLM cell growth in full FBS medium at 9 days after treatment. The proliferative effect of IGF-I was inhibited by a neutralizing antibody to IGF-IRβ (P < 0.0002). The MTT cell proliferation assay was repeated at least three times with comparable results. The absorbance of MTT was expressed as means ± SE of 8 or 16 wells in 96 wells plates for each treatment condition.

Effects of IGF-I on Activation of IGF-IR Pathway in UtLM Cells

To determine whether the IGF-I ligand could activate the IGF-IR and its associated downstream proteins in vivo, UtLM cells were exposed to 100 ng/mL of IGF-I in serum-free medium for 0, 5, 10, 30 and 60 min. Western immunoblots of the cell lysates at different time points were probed with anti-phospho and anti-total IGF-IRβ, IRS-I, Shc, AKT, and MAPKp44/42. UtLM cells exposed to IGF-I were found to have increased protein tyrosine kinase phosphorylation in a time-dependent manner. When UtLM cells were cultured in media containing IGF-I peptide, the phosphorylated IGF-IR and IRS-I protein expression levels were increased within 5 min, reaching maximum levels at 10 min and plateauing by 60 min. However, phosphorylated Shc, MAPKp44/42, and AKT expression levels were increased in 5 min, reached maximum levels at 10 min, thereafter declining toward baseline by 60 min (Figure 6A). The ratio of phosphorylated over total of IGF-IRβ, IRS-I, Shc, and MAPKp44/42 followed the same pattern, and was significantly increased in UtLM cells treated with IGF-I compared with untreated UtLM cells (P < 0.05; Figure 6B).

Figure 6

Activation Effects of IGF-I on IGF-IR pathway in human uterine leiomyoma cells.

(A,B) Phosphorylated IGF-IRβ was significantly higher 5 min after IGF-I (100ng/mL) treatment and remained increased until 60 min (P < 0.05). Downstream adapter protein IRS-1 followed the same pattern (P < 0.05). Phosphorylation of another adapter protein Shc (P < 0.05) and downstream effector protein AKT (P < 0.05) and MAPKp44/42 (P < 0.05) were significantly increased at 5 min and peaked at 10 min when the cells were treated with IGF-I, then fell back to basal levels at 60 min. (C) The activation of IGF-IRβ, Shc, and MAPKp44/42 induced by IGF-I was neutralized partially (Shc: about 40% and MAPKp44/42: about 80%, P < 0.05) when the cells were incubated with anti-IGF-IRB antibody before IGF-I treatment. However, the induction of phosphorylated IRS-I at 10′, 30′, and 60′, and phosphorylated AKT at 5′, 10′, 30′ and 60′, were not significantly decreased. The western blot analysis was done at least three times in three independent in vitro experiments with comparable results. The phospho/total ratio data was expressed as means ± SE of three replicates.

Two hours of blocking with anti-IGF-IRβ resulted in a significant reduction in phosphorylated IGF-IRβ, Shc, and MAPKp44/42 (P < 0.05) compared with UtLM cells in medium containing IGF-I without the neutralization antibody (Figure 6A). Densitometric scanning showed that IGF-IRβ level in anti-IGF-IRβ treated UtLM cells was approximately 20% of that seen in UtLM cells without the antibody treatment. There also was a reduction of approximately 40% in phosphorylated Shc and 80% in phosphorylated MAP kinase levels respectively (Figure 6C). However, the inhibitory effect on expression of phosphorylated IRS-I was not statistically significant at the 10-, 30-, and 60-min time points. In addition, the expression of phosphorylated AKT was not altered significantly at any time point (Figure 6C), which might be due to IGF-IR activation being blocked partially by the neutralization antibody, or the activation of IRS-1 and AKT occurring through activation of the IR in UtLM cells after treatment with IGF-I. These data indicate that the MAP kinase pathway was activated through phosphorylation of IGF-IRβ and its adaptor protein Shc when stimulated with IGF-I at 100 ng/mL.

Discussion

Despite the major impact on gynecological morbidity, the etiology of uterine leiomyomas remains poorly understood (17). Uterine leiomyomas are characterized by changes in cell proliferation and differentiation, and it appears that multiple growth factors are probably important in the pathogenesis of these tumors (1,6–8,18). Growth factors mediate diverse biologic responses through control of cellular proliferation, differentiation, migration, and metabolism by binding to and activating cell-surface receptors that have intrinsic protein kinase activity. To date, about 60 receptor tyrosine kinases (RTKs), which belong to about 16 different receptor gene families, have been identified.

In this study, leiomyoma and patient-matched myometrial samples from ten women were examined for expression of 17 activated growth factor receptor tyrosine kinases. We found that 15 out of the 17 activated RTK receptors evaluated were highly expressed in tumor compared with myometrial samples, and many of these receptors belonged to the EGF, IGF-I, FGF, HGF, and PDGF growth factor gene families, which are important in cell proliferation and differentiation. To our knowledge, this is the first study of growth factor RTK expression profiles in human uterine leiomyoma and matched myometrial tissues.

Several studies have shown that growth factors and their receptor-mediated signaling pathways are important in uterine leiomyoma growth. One such growth factor, EGF, is mitogenic (19) and is expressed more in leiomyomas than in myometrial tissue during the lutereal phase (20). The EGF receptor in leiomyomas is reported to be more sensitive to regulation by sex steroids than those in the myometrium (8,21,22). The growth factor bFGF also induces proliferation of smooth muscle cells in both leiomyomas and myometrial tissue (23). The enhanced growth of leiomyomas may be due partially to the presence of large quantities of bFGF stored in the extracellular matrix (ECM) of these tumors (24). The expression of FGF receptor protein was also reported to be more intense in leiomyomas than in the myometrium (25). Another potent mitogen for smooth muscle cells is PDGF, its mRNA is expressed in both leiomyomas and in myometrium, and its receptor sites per cell are seen more in leiomyomas than in the myometrium (26,27). However, it appears that PDGF does not act alone, but acts in concert with other growth factors such as TGF-β, EGF, and the IGFs. For example, low amounts of TGF-β stimulate autocrine PDGF secretion and promotes the synthesis of PDGF receptors (28). When myometrial cells are treated with both PDGF and EGF, there is a synergistic decrease in DNA synthesis, whereas treatment of leiomyoma cells with both factors results in an additive increase in DNA synthesis (26). Insulin and PDGF also exert an additive effect upon DNA synthesis in leiomyoma and myometrial cells. The mRNA expression level of IGF-I was reported higher in leiomyomas than in the myometrium (26,29,30), although insulin and its receptor are not highly expressed in leiomyoma tissue (31). The levels of IGF-I receptors in leiomyomas have also been reported to exceed those of the myometrium (32–34), which suggests that IGF-I and the IGF-IR signaling pathway may be of major significance in the growth of uterine leiomyomas. There are limited studies done on the role of the hematopoietic growth factors, M-CSF-R, Fit-3, SCF-R, and MSP-R on uterine leiomyoma growth and development, although these RTKs were highly expressed in the leiomyomas compared with myometrial tissues in our RTK array studies. These receptors might be worthwhile studying in the future.

The upregulation of different families of phosphorylated growth factor RTK proteins in the leiomyoma samples found in this study further indicates that there are multiple growth factors that might be important in the pathogenesis and growth of fibroids. Different growth factors could play a role at different stages of the disease. Many of the growth factors may interact, sometimes resulting in a synergistic effect (8). The signal specificity may be defined partially by a combinatorial control. Every RTK recruits and activates a unique set of signaling proteins via its own tyrosine autophosphorylation sites and by means of tyrosine phosphorylation sites on closely associated docking or adaptor proteins. The combinatorial recruitment of a particular complement of signaling proteins from a common preexisting pool of signaling cascades is one mechanism for control of signal specificity (35). This process is regulated further by differential recruitment of stimulatory and inhibitory proteins by the different receptors and downstream effector proteins leading to fine tuning of cellular responses. Signaling pathways activated by RTKs are interconnected with other signaling pathways via protein networks that are subjected to multiple positive and negative feedback mechanisms (35).

Among these highly expressed RTKs in leiomyoma tissues, we observed that the IGF-IR exhibited increased phosphorylation in leiomyomas compared with myometrial samples in this study, and, in earlier studies, we have found the receptor protein to be overexpressed in leiomyomas compared with myometrium (5). IGF-IR affects cell mitogenesis and survival by binding of its ligand, IGF-I, and by activation of downstream effector proteins. Upregulation of IGF-I and/or the IGF-IR could increase fibroid growth and/or survival through its mitogenic and/or anti-apoptotic effects (6,18). IGF-IR is a major receptor tyrosine kinase protein, which appears to be pivotal to the adequate function of other growth factors. Some researchers have reported that overexpression of EGF, PDGF, and insulin receptor is not sufficient for ligand dependent growth unless a functional IGF-IR is present (12,36). On the other hand, overexpression of the IGF-IR renders mouse embryo-derived fibroblasts capable of growing in the presence of IGF-I only, without activation of PDGF and EGF receptors. These findings would suggest a critical role of the IGF-IR in the mitogenic action of other growth factors (13,36). In keeping with this concept and based on increased immunolocalization of IGF-IR in leiomyoma compared with myometrial tissue observed in our earlier studies (5,18), and increased expression levels of phosphorylated IGF-IR in the leiomyoma samples indicated by the RTKs array, we focused on IGF-I and IGF-IR pathway expression and activation in fibroids in this study.

IGF-I, the product of an estrogen-regulated gene, mediates the biologic effectors of growth hormone in many tissues. It exerts its mitogenic action by increasing DNA synthesis, accelerating the progression of the cell cycle from G- to S-phase, and inhibiting apoptosis (6). Several studies have indicated increased expression of IGF-I mRNA and higher tissue concentrations of IGF-I protein in leiomyomas versus corresponding myometrium (37–39). Our previous study also found that IGF-I peptide immunolocalized to the leiomyoma cells and the fibroblasts in the bands of intervening connective tissue comprising the extracellular matrix in some leiomyomas; the latter was not seen in the myometrial samples. In addition, a significant increase in the levels of IGF-IRβ in leiomyomas was noted (5). These data support a possible autocrine or local paracrine mechanism for IGF-I induced growth in leiomyomas. Other data in rats have shown that IGF-I acts as an autocrine growth factor in the regulation of normal growth in the myometrium, and dysregulation of IGF-I signaling could contribute to the neoplastic growth of uterine leiomyomas (9). Our findings of up-regulation of the IGF-I/IGF-IR pathway confirm what has been found in previous studies and demonstrate that higher expression levels of IGF-I and IGF-IRβ are present in leiomyomas compared with myometrial tissue. In addition, our study has demonstrated significantly higher expression levels of phosphorylated IGF-IRβ and its downstream adaptor/effector proteins, Shc, Grb2, and MAPKp44/42. These data support the involvement of the IGF-I/IGF-IR pathway in fibroid growth and development.

The best characterized signaling pathways activated by the IGF-IR are the MAP kinase and the PI3 kinase pathways (40). Tyrosine-phosphorylated IRS-1 and Shc bind different adaptor/effector proteins inducing multiple signaling cascades, among them several interconnecting pathways controlling cell survival and proliferation. Activation of the Shc/Ras/MAP kinase pathway leads to transcriptional responses associated with mitogenesis and cell proliferation. The critical survival pathway activated by IGF-I stems from IRS-1. IRS-1 recruits and stimulates PI3K, which then transmits signals to the serine/threonine kinase AKT. Activated AKT phosphorylates and blocks a variety of proapoptotic proteins. Furthermore, AKT induces the expression of the anti-apoptotic protein Bcl-2 (41,42). In this study, there was minimal expression of PI3K, and phospho- and total IRS-1 in both tumor and myometrial tissues. There was slightly higher expression of phosphorylated AKT in a few tumor samples compared with matched myometrium by immunohistochemical analysis. However, there were no significant differences in total and phosphorylated AKT expression levels observed between tumor and myometrial tissue from confirmative western blotting. This is consistent with our previous findings that a higher rate of cell proliferation appears to play the predominant role in uterine leiomyoma growth, and that neither prolonged cell survival, nor loss of expression of apoptosis-inducing proteins, or increased apoptosis, were likely to be significant mechanisms of uterine leiomyoma cell growth (18). However, another group found higher expression of phospho-AKT in leiomyomas than in myometrial tissue (43,44), which may be due to differences in sampling and/or assays used to determine phosphorylated AKT or total AKT levels. We used immunohistochemistry, western blot analysis, and in vitro studies to evaluate the expression levels of total and phosphorylated AKT in leiomyoma and myometrial tissue and in leiomyoma cells. In our studies, we compared the ratio of phosphorylated AKT over total AKT, which was not done in one study where a difference in expression of phosphorylated and total AKT was noted in leiomyoma versus myometrial tissues (43).

In vitro, IGF-I is mitogenic for a variety of cells including fibroblasts, smooth muscle cells, and leiomyoma cells (11,45,46). In this study, we found that IGF-I treatment resulted in significantly enhanced proliferation of UtLM cells treated with IGF-I peptide compared with non-treated UtLM cells. Interestingly, exogenous IGF-I could compensate for steroid hormones, and possibly growth factor peptide effects, which had been reduced or removed from serum after charcoal/dextran treatment by restoring UtLM cell growth to the level of UtLM cells under full serum culturing conditions. IGF-I treatment also increased phosphorylated IGF-IRβ expression in UtLM cells and facilitated activation of the IGF-IR signaling cascade and the downstream effector proteins Shc and MAPKp44/42, which correlated positively with UtLM cell maintenance and proliferation. Taken together, our results further indicate that activation of IGF-IR reduced the requirement for hormones and other growth factors, and was necessary for UtLM cells to obtain optimal growth in vitro, which is consistent with other groups’ findings of a central role of the IGF-IR in the mitogenic action of the IGF-I peptide and other growth factors (13,36). A neutralizing antibody against IGF-IRβ inhibited IGF-I-induced stimulation of UtLM cell proliferation and partially blocked IGF-I-induced activation of the IGF-IR pathway, which is consistent with some reports that cells in monolayer culture are only partially sensitive to the inhibition of IGF-IR when IGF-IR is blocked (42). The in vitro data further support the involvement of the IGF-IR and MAPK pathways in orchestrating uterine leiomyoma growth.

In contrast to our in vivo findings, the IGF-I peptide increased the phosphorylation of IRS-I and AKT in UtLM cells. These differential effects might be due to differences in the biological environments between tissue and cells grown in culture. The activation of IRS-I and AKT was not blocked significantly by IGF-IR neutralization, which might indicate that IRS-I phosphorylation was most likely induced by the IGF-I peptide binding to the insulin receptor.

In summary, we have analyzed the expression profiles of RTKs in leiomyoma and patient-matched myometrial tissue, and identified phosphorylation of IGF-IR and 14 other growth factor RTKs in leiomyoma tissue. We have also found an overexpression of IGF-I and IGF-IRβ, and downstream phosphorylated effector proteins of the IGF-IRβ signaling pathway in leiomyomas compared with myometrial tissue. These data indicate that activation of the IGF-IR/MAPK pathway in fibroids is important in uterine leiomyoma growth as proposed in Figure 7. Exogenously added IGF-I had a mitogenic effect on UtLM cells, and an effect on activation of IGF-IR and its downstream effector proteins in vitro. A neutralizing antibody against IGF-IRβ inhibited IGF-I-induced stimulation of UtLM cell proliferation and the expression of IGF-IR and downstream proteins.

The differential expression of IGF-I and MAPK pathway proteins in patient-matched leiomyoma and myometrial tissues and in UtLM cells before and after stimulation with IGF-I observed in this study suggests a model for an IGF-I induced signaling cascade in leiomyoma development (Figure 7). In this model, IGF-I peptide binds to the IGF-IR to induce tyrosine autophosphorylation of the β subunit of the receptor and phosphorylation of its adaptor protein Shc. Phosphorylated Shc then associates with Grb2-mSOS complex to activate p21/Ras, which leads to transcriptional activation of genes involved in proliferation through the Ras/Raf/MAPK pathway. IGF-I and IGF-IRβ complex also autophosphorylates its docking protein, IRS-I, which in turn activates the survival PI3K/AKT pathway. In our in vivo study, however, this pathway does not appear to play a major role in the pathogenesis of leiomyomas. IRS-I may also recruit Grb2, but the Shc-Grb2 pathway seems to be a predominant activator of p21/Ras in IGF-IR signaling in UtLM cells and in the pathogenesis of uterine leiomyomas.

Figure 7

Proposed pathway.

In this model, IGF-I peptide binds to IGF-IR to induce tyrosine autophosphorylation and phosphorylation of its adaptor protein Shc. Phosphorylated Shc then is associated with the Grb2-mSOS complex to activate p21/Ras, which leads to proliferation by activation of the Ras/Raf/MAPK pathway. The IGF-I and IGF-IR complex also autophosphorylates its docking protein IRS-I, which, in turn, activates the survival PI3K/AKT pathway. However, on the basis of our studies, this pathway appears not to be playing a major role in leiomyoma growth. IRS-I also may recruit Grb2, but it appears that the Shc-Grb2 pathway is the predominant activator of p21/Ras in IGF-IR signaling in UtLM cells and uterine leiomyoma tissue.

The findings in this study may indicate new anti-tumor targets for noninvasive treatment of uterine leiomyomas. A variety of approaches have been used in preclinical studies to inhibit IGF-IR signaling, including dominant negative mutants, kinase-defective mutants, antisense oligonucleotides, IGFBPs, soluble IGFR antagonistic and/or neutralizing antibodies, and small-molecule kinase inhibitors. Antagonistic antibodies and TK inhibitors are probably the most clinically viable options to date (47,48).

We conclude that upregulation of multiple RTKs and activation of the IGF-I/IGF-IR pathway play an important role in uterine leiomyoma growth. The results from this study potentially may provide non-invasive therapeutic interventions for clinical cases of uterine fibroids.