|Home | About | Journals | Submit | Contact Us | Français|
Endometrial cancer (EMC) is the most common gynecological malignancy. The etiology and the cell types that are conducive to EMC are not completely understood, provoking further studies. Our objective was to determine whether deletion of Pten specifically in the uterine stroma and myometrium induces cancer or manifests different phenotypes.
PtenAmhr2(d/d) mice with conditional deletion of Pten in the mouse uterine stroma and myometrium, but not in the epithelium, were generated by mating floxed Pten mice and anti-Mullerian hormone type 2 receptor (Amhr2)-Cre mice. The phenotypes were compared between Ptenf/f and PtenAmhr2(d/d) uteri.
We show that conditional deletion of Pten in the mouse uterine stroma and myometrium, but not in the epithelium, fails to generate EMC even at the age of 5 months. Surprisingly Pten deletion by Amhr2-Cre transformed a large number of myometrial cells into adipocytes with lipid accumulation, possibly a result of increased levels of SREBP1 and PPARγ which regulate adipose differentiation.
These results provide evidence that deletion of Pten specifically in the stroma and myometrium does not result in EMC in female mice examined up to 5 months of age but alters the myocytes to adipocytes and mimics histologic similarities with lipoleiomyomas in humans, raising the possibility of using this mouse model to further explore the cause of the disease.
Endometrial cancer (EMC) is the most common gynecological malignancy. In fact, 6% of U.S. female cancer patients exhibit EMC (1). According to the American Cancer Society, nearly 42,000 new cases of EMC were diagnosed and 7,800 patients died from the disease in 2009 (1). There are several treatment options effective for the early stages of EMC, such as hysterectomy and hormonal therapy; however, only limited options remain if the cancer metastasizes (2). The etiology and the cell types that give rise to EMC are not completely understood and further studies are warranted to develop treatment options.
Animal models that spontaneously develop cancer are powerful tools for studying the mechanism of cancer initiation and progression and for developing treatment strategies. We previously reported a mouse model of endometrial carcinoma in which the floxed phosphatase and tensin homologue (Pten) gene was deleted by Cre expression under the control of a progesterone receptor (PR) promoter (3). PTEN is a tumor suppressor with the highest frequency of mutation in human EMCs (2). In our mouse model, uterine deletion of Pten resulted in severe complex hyperplasia of the uterine epithelium by three weeks of age with carcinoma occurring by one month. More importantly, our mouse model of EMC resembles many aspects of type I human EMCs (3).
Conditional deletion of Pten in the mouse uterus using PR-Cre results in the development of cancer in the epithelium. Given that PR is expressed in the epithelium, stroma, and myometrium, the question remains whether Pten deletion in the uterine stroma and/or myometrium also contributes to EMC (4). To address this question, we used anti-Mullerian hormone type 2 receptor-Cre (Amhr2-Cre) mice to specifically delete Pten in the uterine stroma and myometrium, since Amhr2 is expressed in these tissue types, but not in the epithelium (5,6). We found that PtenloxP/loxP/Amhr2cre/+ (PtenAmhr2(d/d)) mice did not develop cancer in the uterus even at 150 days of age. Remarkably, many myometrial cells in these mice accumulated lipids and these uteri expressed markers of white adipose tissue, possibly due to the observed increases in levels of SREBP1 (sterol regulatory element binding protein 1) and PPARγ (peroxisome proliferator-activated receptor γ) that are known to regulate adipose differentiation (7–10). Further investigation as to how myometrial cells transform into adipocytes in PtenAmhr2(d/d) mice may provide valuable information relevant to lipoleiomyomas in humans.
All mice were housed in the Cincinnati Children's Hospital Medical Center Animal Care Facility according to NIH and institutional guidelines for laboratory animals. While PtenloxP/loxP mice (stock number 004597, 129S4/SvJae/BALB/cAnNTac) were obtained from the Jackson Laboratory, Amhr2-Cre mice were obtained from Richard Behringer (MD Anderson Medical Center) and PR-Cre mice were obtained Francesco DeMayo and John Lydon (Baylor College of Medicine). PCR analysis of tail genomic DNA determined the genotypes of mutant mice.
Paraformaldehyde-fixed frozen sections were hybridized with 35S-labeled cRNA probes as described (3).
LacZ staining was performed as previously described (11). In brief, tissues were embedded in OCT after fixation in 0.2% paraformaldehyde and infused in 30% sucrose at 4°C. Frozen sections were stained with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside overnight at 37°C. Sections were counter-stained with eosin.
Immunohistochemistry was performed as previously described (12). In brief, tissues were fixed in PROTOCOL Safefix II (Thermo Fisher) and embedded in paraffin. Uterine sections (6 μm) were subjected to immunostaining using antibodies to cytokeratin 8 (CK8; Developmental Studies Hybridoma Bank), α-smooth muscle actin (αSMA; Abcam), or PPARγ (Santa Cruz). After deparaffinization and hydration, sections were subjected to antigen retrieval by autoclaving in 10 mM sodium citrate solution (pH = 6) for 10 min. A diaminobenzidine (DAB) kit (Invitrogen) was used to visualize antigens. Sections were counter-stained with hematoxylin.
Frozen sections (12 μm) were fixed in 10% formalin, treated with propylene glycol, and stained with Sudan black B solution (0.7% in propylene glycol). After staining, sections were rinsed with 85% propylene glycol followed by washing in distilled water (13,14).
Frozen sections (12 μm) were fixed in 10% formalin solution and stained with Oil O Red solution (0.5% in isopropanol). After staining, sections were rinsed with distilled water and mounted.
RNA samples were prepared and analyzed as previously described (15). In brief, total RNA was extracted with RNAeasy Protect Mini kit (Qiagen) according to the manufacturer's protocol. Samples were obtained from three Ptenf/f or PtenAmhr2(d/d) uteri. Quantitative RT-PCRs using Taqman probes (Applied BioSystems) were performed with primer sets specific for Adfp (adipose differentiation-related protein), Lpl (lipoprotein lipase), and Fabp4 (fatty acid binding protein 4). A probe set for 18S rRNA was used as the normalization standard. The PCR and relative quantifications were performed using 25 ng of cDNA per reaction in a real-time PCR system. Relative quantification data from quantitative PCR analysis were statistically analyzed using Student's t-tests.
Tissue samples were prepared as previously described (12). After measuring protein concentrations, supernatants were mixed with SDS-PAGE sample buffer and boiled for 5 min. Samples were run on 10% SDS-PAGE gels and transferred onto PVDF membranes. Membranes were blocked with 10% milk or 5% BSA in TBST and probed with antibodies to pAKT (Cell Signaling), AKT (Cell Signaling), SREBP1 (Santa Cruz), or ACTIN (Santa Cruz) overnight at 4°C. After washing, blots were incubated in peroxidase-conjugated donkey anti-goat or donkey anti-rabbit IgG (Jackson Immuno Research Laboratories, Inc.). All signals were detected using chemiluminescent reagents (GE Healthcare). ACTIN and AKT served as loading controls.
Our first objective was to study whether deletion of Pten specifically in the uterine stroma and myometrium induces cancer, since Pten is also expressed in the stroma and myometrium (Fig. 1A). For this study, we used Amhr2-Cre transgenic mice to delete Pten in the stroma and myometrium (5,6). Using Rosa26R mice, we confirmed that Cre induced by the Amhr2 promoter successfully recombined loxP sites only in the stroma and myometrium. Recombination did not occur in the epithelium, even at 60 days of age, while PR-Cre very rapidly recombines loxP sites in the epithelium, stroma, and myometrium (Fig. 1B). Of note, the recombination was also observed in the ovary and oviduct. We examined uterine histology at 150 days of age in PtenloxP/loxP/Amhr2cre/+ (PtenAmhr2(d/d)) mice (n=9). As shown in Figure 2A, PtenAmhr2(d/d) uteri had only uncommon foci of complex atypical hyperplasia (CAH), but not EMC, while PtenPr(d/d) females developed EMC (Fig. 2D). Ovarian histology in these mice appeared normal (n=7).
Remarkably, many myometrial cells in PtenAmhr2(d/d) uteri showed histological characteristics of adipocytes (n=9), and staining with Sudan Black B and Oil O Red showed accumulation of neutral lipids in those cells (Table 1, Fig. 2B & C). This phenotype was not evident in mice in which uterine Pten was deleted by PR-Cre (PtenPr(d/d)) even at 150 days of age (n=4; Fig. 2D). To assess whether myometrial adipogenesis was progressive, we examined lipid accumulation by Sudan Black B staining in uteri of PtenAmhr2 (d/d) at the age of 30 and 60 days. Notably, Amhr2 expression is observed in the myometrium as early as 14 days of age (6). While no positive staining for lipid accumulation was noted in the myometrium of PtenAmhr2 (d/d) females at 30 days of age (n=6, Fig 3A), foci of positive staining were evident in myometrium of PtenAmhr2 (d/d) females by 60 days of age (n=5, Fig 3B). The accumulation of lipid droplets was more restricted in uteri at the age of 60 days than seen at 150 days. These results suggest that myometrial adipogenesis started between 30 to 60 days of age in PtenAmhr2 (d/d) females.
To further confirm that myometrial cells transformed into adipocytes, we analyzed the expression levels of Adfp (adipose differentiation-related protein), Lpl (lipoprotein lipase), and Fabp4 (fatty acid binding protein 4), known markers of adipogenesis by quantitative RT-PCR (16–18). Expression levels of Lpl and Fabp4 were significantly increased in PtenAmhr2(d/d) uteri (Fig. 4A). In contrast, levels of Adfp were similar between PtenAmhr2(d/d) and Ptenf/f uteri, indicating that Adfp did not participate at the formation of intracellular lipid droplets (Fig. 4A). These data indicate that myometrial Pten deletion transforms myometrial cells into adipocytes.
Akt has been shown to upregulate lipogenesis in adipose and muscle tissues (reviewed in (19)) and there is also evidence that activation of Akt increases the level of a transcription factor SREBP1 (sterol regulatory element binding protein 1) in osteosarcoma cells (20). SREBP1 activates another transcription factor, PPARγ, and induces adipose differentiation in 3T3-L1 cells (7–10). Since pAKT levels are upregulated in PtenPr(d/d) uteri with the development of EMC (3), we also examined the levels of pAKT as well as the status of SREBP1 and PPARγ in PtenAmhr2(d/d) uteri.
We found that the levels of pAKT and SREBP1 are upregulated in PtenAmhr2(d/d) uteri compared to Ptenf/f uteri as evident from Western blot analysis (Fig. 4B). In addition, PPARγ immunostaining showed an increase in myometrial cells with distinct nuclear localization of PPARγ in PtenAmhr2(d/d) mice compared to Ptenf/f mice (Fig. 4C). Notably, PPARγ nuclear staining was mainly observed in the longitudinal muscle layer, where adipogenesis occurred after Pten deletion. These results suggest that the transformation from myometrial cells to adipocytes by Pten deletion is mediated through the activation of the pAKT-SREBP1-PPARγ pathway (Fig. 4D).
The observation that PtenPr(d/d) mice with conditional deletion of Pten in all major uterine cell types and PtenAmhr2(d/d) mice with conditional deletion of Pten in the uterine stroma and myometrium produce different phenotypes is very interesting. These results establish a new paradigm that deletion of Pten in the stroma/myometrial compartment is not sufficient to drive the development of EMC even by the age of 150 days, although the influence of the stroma in the initiation and development of EMC cannot be ruled out, since foci of CAH were seen. Addressing this issue would require deleting Pten specifically in the uterine epithelium. Unfortunately, our attempts to delete epithelial Pten using Wnt7a-Cre (provided by Richard Behringer) met with failure; PtenWnt7a(d/d) mice died around 10 days of age (data not shown). It was recently reported that epithelial-specific deletion of Lkb1 in the uterus using Sprr2f-Cre produces EMC, suggesting that EMC primarily originates from epithelial cells (21). However, Sprr2f-Cre mice cannot be used in studying PTEN's roles in EMC, since deletion of Pten by Sprr2f-Cre induces brain cancer due to Cre expression in the cerebellum and limits the life span of those mice (personal communication, Diego Castrillon, UT Southwestern Medical Center) (21). In addition, a recent report shows that only the epithelial absence of Pten can induce EMC in an in vivo endometrial regeneration system (22). Collectively, our results and the results of others provide evidence that EMC is of epithelial origin.
Another significant finding of our present study is the transformation of myometrial cells into adipocytes in PtenAmhr2(d/d) mice. This transformation, however, does not occur in mice with Pten conditionally deleted in all major uterine cell types (epithelium, stroma, and myometrium) using PR-Cre. PTEN's role in adipogenesis has previously been reported in other systems. Chondrocyte-specific deletion of Pten leads to the formation of adipose tissues and lipomas along the spine (23,24), hepatocyte-specific deletion of Pten leads to steatohepatitis with hepatocellular carcinomas, and patients with Pten mutations often develop vascular anomalies with ectopic accumulations of fat (25). However, myocytes do not transform into adipocytes when Pten is conditionally deleted in skeletal muscles and cardiomyocytes in mice (26,27). These results may suggest that the transformation occurs only in smooth muscles. In the myometrium, PTEN's role in adipogenesis is specific to the longitudinal muscle layer; the reason for this specificity remains unknown.
In this study, we provide evidence that Pten deletion in the uterine stroma and myometrium activates the pAKT-SREBP1-PPARγ pathway to trigger myometrial adipogenesis (Fig. 4D). Increased levels of SREBP1 and PPARγ were also observed in mice with hepatocyte-specific Pten deletion, resulting in steatohepatitis (24). However, defining the exact mechanism and the role of stroma and smooth muscle in this process would require further investigation. There is evidence that conditional deletion of β-catenin using Amhr2-Cre mice switches myogenesis to adipogenesis in the uterine myometrium of adult mice. Since Amhr2 (also known as Mullerian inhibiting substance type receptor) is expressed in the Mullerian duct mesenchyme before postnatal differentiation of the uterus into distinct tissue types, it was suggested that deletion of β-catenin in the mesenchymal tissue resulted in the switching of myogenesis to adipogenesis in the adult myometrium (28). This perhaps explains the transformation of myocytes to adipocytes in mice conditionally deleted of Pten by Cre expression driven by the Amhr2 promoter as opposed to Cre driven by PR whose expression becomes prominent from postnatal day 10. It is interesting to note that impairment of both canonical Wnt signaling and PTEN signaling have similar effects in shifting myogenesis to adipogenesis, although deficiency of these two pathways has opposing effects on tumorigenesis. Further investigation is warranted to better understand this complex process.
This study was supported in parts by the Concern Foundation, Ohio Cancer Research Associates, and the Perinatal Institute, CCHMC (T Daikoku), and NIH grants HD12304 and PO1-CA-77839 (SK Dey). We thank Erin L. Adams for editing the manuscript. We are grateful to Richard M. Behringer (MD Anderson Cancer Center, Houston, TX) and Francesco DeMayo and John B. Lydon (Baylor College of Medicine, Houston, TX) for providing the Amhr2-Cre and PR-Cre mice, respectively.
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.
Conflict of interest statement The authors declare that there are no conflicts of interest.