Distinct Function of Estrogen Receptor in Smooth Muscle and ...

ORIGINAL

RESEARCH

Distinct Function of Estrogen Receptor ␣ in Smooth Muscle and Fibroblast Cells in Prostate Development Spencer Vitkus, Chiuan-Ren Yeh, Hsiu-Hsia Lin, Iawen Hsu, Jiangzhou Yu, Ming Chen, and Shuyuan Yeh Departments of Urology and Pathology, University of Rochester Medical Center Rochester, Rochester, New York 14642

Estrogen signaling, through estrogen receptor (ER)␣, has been shown to cause hypertrophy in the prostate. Our recent report has shown that epithelial ER␣ knockout (KO) will not affect the normal prostate development or homeostasis. However, it remains unclear whether ER␣ in different types of stromal cells has distinct roles in prostate development. This study proposed to elucidate how KO of ER␣ in the stromal smooth muscle or fibroblast cells may interrupt cross talk between prostate stromal and epithelial cells. Smooth muscle ER␣KO (smER␣KO) mice showed decreased glandular infolding with the proximal area exhibiting a significant decrease. Fibroblast ER␣KO mouse prostates did not exhibit this phenotype but showed a decrease in the number of ductal tips. Additionally, the amount of collagen observed in the basement membrane was reduced in smER␣KO prostates. Interestingly, these phenotypes were found to be mutually exclusive among smER␣KO or fibroblast ER␣KO mice. Compound KO of ER␣ in both fibroblast and smooth muscle showed combined phenotypes from each of the single KO. Further mechanistic studies showed that IGF-I and epidermal growth factor were down-regulated in prostate smooth muscle PS-1 cells lacking ER␣. Together, our results indicate the distinct functions of fibroblast vs. smER␣ in prostate development. (Molecular Endocrinology 27: 38 – 49, 2013)

strogen actions in the prostate can be mediated via two distinct receptors, estrogen receptor (ER)␣ or ER␤ (1). These two receptors, although displaying similar structural homology and ligand binding specificity, control two distinct pathways in the prostate. ER␣ appears to be involved in cellular proliferation signaling, whereas ER␤ is implicated in antigrowth, apoptotic, signaling (2). In addition to having different functions, these two receptors are localized to different prostate cell types. In mice, ER␣ is found in the mesenchymal cells, whereas ER␤ is found predominantly in the epithelial cells (3–5). A growing body of evidence indicates that estrogen signaling is important in prostate ductal morphogenesis (6), extracellular matrix composition (7) and disorders, such as prostate cancer and benign prostatic hyperplasia (BPH) (8 –12). Pregnant female mice treated with low dose, but

E

above physiological, levels of estrogen gave birth to male offspring that displayed hyperplastic phenotypes at adulthood (13). This indicates that one or both of the ERs are critical for prostate development. Prins et al. (5) showed that the hyperplastic phenotypes described earlier were not observed in male ER␣ knockout (KO) mice from estrogenized mothers. In ER␤KO mice, hyperplasia was still seen, confirming that ER␣, but not ER␤, was the primary receptor involved in aberrant cellular growth. Interestingly, because hyperplasia was seen in the epithelial cell population, but ER␣ is expressed in the stromal compartment and a subset of basal cells, cross talk between the stromal and epithelial compartments occurs, and indeed, this phenomenon has been shown on numerous occasions (14 –17). Disruption of this cross talk has been tested in tissue recombination models but has not

ISSN Print 0888-8809 ISSN Online 1944-9917 Printed in U.S.A. Copyright © 2013 by The Endocrine Society doi: 10.1210/me.2012-1212 Received June 20, 2012. Accepted October 29, 2012. First Published Online November 30, 2012

Abbreviations: AR, Androgen receptor; BMP4, bone morphogenic protein 4; BPH, benign prostatic hyperplasia; CCL2, (C-C motif) ligand 2; CM, conditioned media; dER␣KO, double-cre ER ␣ KO; E2, 17␤-estradiol; EGF, epidermal growth factor; ER, estrogen receptor; FBS, fetal bovine serum; fER␣KO, fibroblast ER␣KO; KO, knockout; MMP9, matrix metalloproteinase 9; PIN, prostatic intraepithelial neoplasia; Q-PCR, quantitative-PCR; smER␣, smooth muscle ER␣; TGF␤R2, TGF␤ receptor II; VP, ventral prostate; Wt, wild type.

38

mend.endojournals.org

Mol Endocrinol, January 2013, 27(1):38 – 49


been truly validated in an in vivo model that recapitulates the intact prostate microenvironment. Two previous studies using total ER␣KO mice have shown a significant 1.7- to 2.0-fold increase in testosterone levels at adulthood (6, 18). After castration, the prostate shrinks in size (19). This diminishment in prostate size is reversible after the addition of testosterone or dihydrotestosterone (20, 21). Combined, these data exhibit the potent mitogenic effects of testosterone in the prostate. Although testosterone levels were increased, our group was still able to observe defects in prostate development in ER␣KO male mice, indicating a direct ER␣ role in prostate development. Previous work in our lab has looked at the effect of total ER␣KO using the ACTB promoter (6). These mice display distinct developmental defects that cannot be reproduced in epithelial ER␣KO mice; thus, stromal ER␣ and not epithelial ER␣ must be responsible for the observed developmental phenotypes. Using fibroblast ER␣KO (fER␣KO) mice, we found that many of the reported phenotypes could be accounted for (22). However, estrogen and ER␣ have been reported to play varying roles in the smooth muscle of uterus (23), endothelium (24 –26), and prostate (6, 27, 28) and to have a particular role in collagen deposition and formation in the prostate (7), yet little is known about the role of smooth muscle ER␣ (smER␣) in prostate development. In the current study, we first looked at the effect of smER␣KO on the ventral prostate (VP) development. We used Tgln cre (29), which was shown to have high efficiency in the VP (30), to create smER␣KO mice. We compared this phenotype with the previously published fER␣KO mice phenotype (22) as well as to mice with ER␣KO in both fibroblasts and smooth muscle cells (double-cre ER ␣ KO, dER␣KO). We found that these dER␣KO mice display an aggregate phenotype of both fibroblast and smooth muscle KO mice. Previous studies testing the importance of ER␣ in the prostate development or pathogenesis used tissue recombination in a semi in vivo fashion. In these models, ER␣ is manipulated in the mesenchyme and then incubated with epithelial cells to see whether prostrate regrowth is possible (31). However, this system uses stromal cells from seminal vesicles, not from prostate, and the coinoculation of these stromal and epithelial cells cannot result in the formation of the normal, full structure of the prostate glands. In addition, these models do not take into account the full prostate microenvironment, which is composed of two main compartments, the epithelial and stromal. The epithelial compartment contains luminal, secretory, and basal cells, whereas the stroma is comprised of fibroblasts, smooth muscle cells, and, to a lesser extent, endo-


39

thelial cells, and infiltrating immune cells (32–35). Tissue recombination distills this complex microenvironment system down to a system involving fewer cell types. In our model, we demonstrate that different cell types invoke different ER␣ actions, providing a more powerful model to mimic the in vivo prostate microenvironment with which to further study the ER␣ role in prostate organogenesis, homeostasis, or pathogenesis.

Materials and Methods Animals Animals were housed in the S wing vivarium at the University of Rochester, School of Medicine. All protocols related to animals were overseen and approved by the animal care and use committee, and all animals were treated in accordance with National Institute of Health guidelines. The Tgln mice were created on an FVB background (29) and backcrossed more than 8 generations to a C57BL/6 background. FSP, and floxed ER␣, mice were created on a C57BL/6 background. The genotypes of above mice were identified using PCR on DNA obtained from tails lysed in buffer (QIAGEN, Valencia, CA) with 0.5 mg/ml proteinase K (Invitrogen, Carlsbad, CA) overnight.

Mouse prostate dissection and histology analysis Mice were euthanized, serum obtained via cardiac puncture, and the whole urogenital tract excised into PBS. Further dissection of specific lobes was achieved under an L2 illumination microscope (Leica, Heerbrugg, Switzerland). The ventral, anterior, and dorsal-lateral prostates were removed to 4% paraformaldehyde for 4 – 6 h depending on size. After fixation, lobes were processed through a gradient of alcohols to xylene and then embedded in paraffin. Sections of 5-␮m thickness were affixed to slides and stained with hematoxylin and eosin. Slides were viewed using an eclipse E800 microscope (Nikon, Melville, NY), and images were taken using Spot Advanced camera software (Diagnostic Instruments, Inc., Sterling Heights, MI). ImageJ (National Institute of Health, Bethesda, MD) was used for image analysis.

Immunohistochemistry and immunofluorescence Slides were dewaxed, rehydrated, and subjected to antigen retrieval in 10 mM sodium citrate buffer (pH 6.0) for 20 min at 96 C in a histowave oven (Thermo Shandon, Pittsburgh, PA). After antigen retrieval, slides were incubated in 3% H2O2 in methanol for 30 min to quench endogenous peroxidase, blocked in a 5% fetal bovine serum (FBS), 5% bovine serum albumin, and 5% nonfat milk solution for 1 h, and the primary antibody applied overnight. Antibodies used were, mouse ER␣ (MC-20, 1:400) and Ki67 (NCL-Ki-67p, 1:1000). The biotinylated secondary antibody (Vector Laboratories, Burlingame, CA) was applied for 1 h, and then slides were incubated with an AvidinBiotin Complex (Vector Laboratories). Development was achieved through use of 3⬘-3⬘-diaminobenzidene (Vectastain), and then Mayer’s hematoxylin was used to counterstain the slides. For immunofluorescence, after the first antibody, a fluorescein isothiocyanate-labeled secondary antibody was applied

40

Vitkus et al.

SMC and fER␣ in Prostate Development

(Invitrogen) to tissue. Tissue slides were then counterstained with 4⬘,6-diamidino-2-phenylindole and mounted.

RNA extraction and real-time PCR TRIzol (Invitrogen) was used per the manufacturer’s instructions with the exception that the quantities were scaled down by half to account for the small size of the prostate. TRIzol in a volume of 500 ␮l was added to each prostate lobe and tissues homogenized using an electric homogenizer. RNA was extracted from the aquous phase, after incubation with phenol: choloroform, using isopropanol. RNA was then washed with 75% ethanol. Real-time PCR was performed using 1 ␮g of RNA, 4 ␮l of 5⫻ iScript reaction mix, 1 ␮l of iScript reverse transcriptase (Bio-Rad Laboratories, Hercules, CA), and enough water to bring the total volume to 20 ␮l. Reverse transcription was completed using the program 95 C for 5 min, 50 C for 45 min, and 70 C for 15 min.

Quantitative-PCR (Q-PCR) of smER␣-regulated gene profile To each well of a 96-well plate, 4 ␮l of a 1:10 dilution of cDNA, 5 ␮l of iQ SYBR Green Supermix reagent (Bio-Rad Laboratories), and 1 ␮l of the appropriate primer were added. Q-PCR was run using an iCycler (Bio-Rad Laboratories). Data were normalized to ␤-actin using the ⌬⌬CT method. Primers used to detect the gene profile of rat smooth muscle PS-1 cells are shown in Table 1.

Branching morphogenesis Prostates were removed en bloc from animals into PBS. After microdissection of individual prostate lobes, each lobe was incubated in 1% collagenase in PBS at 37 C for the following

TABLE 1.


times: VP, 5 min; dorsal-lateral prostate, 10 min; and anterior prostate, 15 min. After collagenase incubation, tissues were removed to a PBS wash and then placed in fresh PBS for further dissection. Stroma was carefully peeled away on a compression slide using fine forceps and a 30-gauge needle tip. Pictures were taken using a Leica MZ1GF microscope with attached camera DFC480 and IM50 Image Manager imaging software (Leica). Ductal tips, branches, and branch points were counted and quantified.

Testosterone ELISA assay Mouse blood was collected by cardiac puncture. The blood was collected into a serum separator tube, spun down, and serum was collected to detect testosterone using an ELISA kit following the manufacturer’s instructions. Briefly, 50 ␮l of serum were added to the supplied microtitration strips, and 100 ␮l of testosterone and 100 ␮l of enzyme conjugate solution were added. Strips were shaken vigorously for 1 h and then washed five times with wash solution. The 3,3⬘,5,5⬘-tetramethylbenzidine chromagen solution was added, the strips shaken for 30 min, and then 100 ␮l of stop solution were added to stop the reaction. The absorbance was read at 450 nM with an ELISA plate reader (BioTek Instruments, Inc., Winooski, VT) and the data analyzed with SoftMax Pro 3.1.1 (Molecular Devices Corp., Sunnyvale, CA).

Cell studies PS-1 and BPH-1 cells were maintained in normal media (DMEM) (Invitrogen) with 10% FBS. Cells were infected using a lentivirus pWPI-blasticidin vector containing flag-mER␣ or vector control plasmid and subsequently selected using 5 ␮g/␮l blasticidin.

Growth factor qPCR primer list Primer target IGF IGF binding protein 5 Platelet-derived growth factor ␣ EGF

␤-Actin Chemokine (C-C motif) ligand 2 Fibroblast growth factor 10 Fibroblast growth factor 7 Bone morphogenic protein 4 Fibroblast growth factor 2 Androgen receptor G protein-coupled ER F: Forward; R, reverse.

F: R: F: R: F: R: F: R: F: R: F: R: F: R: F: R: F: R: F: R: F: R: F: R:

Primer sequence 5⬘-TGTGTACTGTCTTCTTCTGCCTGCA-3⬘ 5⬘-AGCTCCCACAGAACCGCACA-3⬘ 5⬘-ATCTGACCAAGGCCCCTGCC-3⬘ 5⬘-GCAGATGCCACGTTTGCGGC-3⬘ 5⬘-CGACTCAGGTCCAGCGTGGT-3⬘ 5⬘-ACAGAGGCACCCTCTCTTGGCC-3⬘ 5⬘-CCATCCTCAACTTTTCTGGGGCTCA-5⬘ 5⬘-ACACGGGGAAGGCCAGAGAGC-3⬘ 5⬘-GCGTCCACCCGCGAGTACAA-3⬘ 5⬘-TCCATGGCGAACTGGTGGCG-3⬘ 5⬘-ACGTGCTGTCTCAGCCAGATGC-3⬘ 5⬘-GCTTCTTTGGGACACCTGCTGCT-3⬘ 5⬘-CGGGGAGGCATGTGCGAAGC-3⬘ 5⬘-GTCCCGCTGACCTTGCCGTT-3⬘ 5⬘-ACACCCGGGGCACTGCTCTA-3⬘ 5⬘-CAGTTCACGCTCGTGGCCGT-3⬘ 5⬘-GCGGGACTTCGAGGCGACAC-3⬘ 5⬘-ATCCGGGATGACGGCGCTCT-3⬘ 5⬘-AACGGCGGCTTCTTCCTGCG-3⬘ 5⬘-AGTTTGACGTGTGGGTCGCTCT-3⬘ 5⬘-GGCTACACTCGGCCCCCTCA-3⬘ 5⬘-CTGTCCAAACGCATGTCCCCA-3⬘ 5⬘-CGCCGTGCTCTGCACCTTCA-3⬘ 5⬘-GCTTTGGCCAGCGCCAGGTA-3⬘


Conditioned media (CM) and growth of BPH-1 PS-1 ER␣ or vector lentivirus-infected cells were plated in media containing 5% charcoal-stripped FBS for 1 d in a 10-cm2 plate. On the 2nd day, the media were changed, plates were treated with 10 nM 17␤-estradiol (E2), and CM was collected 24 h after treatment. BPH-1 cells were plated 5 ⫻ 103 cells/well in a 24-well plate in 5% charcoal-stripped FBS media and allowed to attach overnight. The next day BPH-1 cells were subjected to CM as d 0. Fresh CM was added to cells every other day. Growth was measured by exposing cells to 5% 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide for 2 h and then dimethylsulfoxide for 1 h. Absorbance was read at 560 nM.

Masson Trichrome staining Visual confirmation of collagen loss was obtained by use of the Accustain Trichrome kit (Sigma Diagnostics, St. Louis, MO). Slides were deparaffinized and then soaked in preheated 55 C Bouins solution for 15 min. Slides were allowed to cool in running tap water for 5 min and then stained in working Weigert’s iron hematoxylin solution for 5 min, rinsed in deionized water, and stained in Biebrich scarlet acid fuschin for 2 min. Slides were rinsed again and then placed in working phosphotungstic/phosphomolibdic acid solution for 5 min, then placed directly into aniline blue for 30 min. Slides were washed in 1% acetic acid for 2 min and then dehydrated and mounted. ImageJ was used for separating out the blue color channel and then quantifying the amount of collagen per area of the basement membrane.

Picrosirius red staining After rehydration, slides were stained in Weigert’s hematoxylin for 8 min. Slides were rinsed and then placed serially through solutions A (2 min), B (60 min), and C (2 min) from the picrosirius red kit (Polysciences, Inc., Warrington, PA), with a wash in between each solution. Slides were dehydrated through xylene and mounted. Red collagen color was quantified using ImageProPlus (Media Cybernetics, San Diego, CA). For each field (total of 15 or more high-powered fields/mouse), the basement membrane was isolated as a region of interest using ImageProPlus (Media Cybernetics). The total area of red collagen fibrils was divided by the area of the basement membrane in that field. The relative amount of collagen was then averaged among the field and this average compared between wild-type (Wt) and smER␣KO mice.

Statistical analysis Student’s two tailed t test with Welsh correction was used to determine significance between mice branching morphogenesis, testosterone levels, and collagen fraction. One way ANOVA was used to analyze growth data, and two way ANOVA with Bonferroni post hoc test was used to analyze the IGF-I/epidermal growth factor (EGF) time course and prostate glandular infolding. GraphPad Prism 5 (GraphPad Software, San Diego, CA) and Microsoft Office 2007 (Microsoft, Redmond, WA) software was used in deriving significance.


41

Results Generation and characterization of smooth muscle-specific ER␣KO mice Floxed ER␣ mice were created using the cre/lox method as described previously (6, 27). Tgln is encoded by the SM22␣ gene in smooth muscle tissue (36). Combining flox ER␣ mice with cre recombinase linked to the Tgln promoter, we were able to create smER␣KO mice (Fig. 1A). For experimental purposes, ER␣ F/F mice or ER␣ F/⫹ littermates were used as Wt controls. Genotyping results from a smER␣KO mouse and Wt littermate are demonstrated in Fig. 1B. To control for any aggressionrelated increase in testosterone levels, mice were separated for 1 wk before serum was collected and analyzed for circulating testosterone levels. These levels were not significantly altered between smER␣KO mice and Wt controls (Fig. 1C), indicating that any change in phenotype is not related to variation in testosterone levels. Further comparison of other organs, including bladder, intestine, stomach, aorta, and testes, and heart function revealed no significant difference (Supplemental Fig. 1, published on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org). smER␣ plays a key role in regulating prostate glandular infolding The prostate is a complex structure that, in the mouse, matures and becomes quiescent by sexual maturity (37). To examine the phenotype changes after complete prostate maturation, mice were euthanized at 12 wk of age. Along with proximal to distal growth, the prostate epithelial cells also form infolding within each duct, thereby increasing surface area (38). If prostate growth is affected by stromal ER␣, then it may also stand to reason that epithelial infolding is affected through stromal-epithelial interactions. To evaluate the effect of ER␣KO in smooth muscle and fibroblast cells on prostate development, prostate histololgy was evaluated. In smER␣KO mice, prostate morphology is changed in a spatial manner with the proximal area of the gland exhibiting less infolding compared with Wt mice (Fig. 2, A and A-1). This decrease was quantified as the number of folds per square millimeter of epithelial circumference (representative fields used for counting are displayed in Supplemental Fig. 2). This change was found to be significant. fER␣KO mice do not display this phenotype and appear to have glandular structure similar to that of Wt mice (Fig. 2A). The prostate develops in a proximal to distal fashion, and it is possible that during the initial growth phase, smER␣ regulates infolding. However, in later growth phases, the distal area of the smER␣KO prostate is able to catch up to

42

Vitkus et al.



FIG. 1. Breeding and characterization of smER␣KO mice. A, Tgln Cre male mice were mated with floxed ER␣ mice. The resultant Tgln heterozygous floxed ER␣ mice were then mated again to floxed ER␣ mice to produce smER␣KO mice. B, Genotyping of smER␣KO. Lane 1 demonstrated the Tgln cre band at 233 bp and the homozygous flox band at 881 bp. Wt mice (lane 2) showed either a single floxed ER␣ band or one floxed ER␣ band and one Wt ER␣ band at 741 bp (not shown). C, Testosterone levels of Wt and smER␣KO male mice. The sera were collected from smER␣KO and age-matched Wt littermates of 12 wk old, and testosterone levels were detected using ELISA. Testosterone levels remain unchanged between Wt and smER␣KO mice.

its Wt counterpart due to loss of ER␣ and thinning of the smooth muscle cells toward the distal end of the prostate during this time. Collagen and basement membrane show disregulation in smER␣KO mouse prostates Because evidence suggests that glandular development can be influenced via the extracellular matrix and basement membrane (7), we decided to stain mouse tissues with picrosirius red to identify collagen deposition. The basement membrane is composed of a host of proteins, one of the major ones being collagen type IV, although other collagens are present, such as III and VII (39). In the smER␣KO mouse model, the VP shows a significant reduction in basement membrane thickness, quantified as the total amount of collagen present over the basement membrane thickness per field of picrosirius red-stained mouse VPs (Fig. 2, B and B-1). Interestingly, the fER␣KO mice do not display this phenotype and show no significant difference when collagen amount is compared with Wt mice. Along with our previous results, this indicates that ER␣ exhibits its influences via a spatial and cell typespecific manner. Prostate bud formation is reduced in fER␣KO mice but not in smER␣KO mice The prostate growth is a tightly regulated process that involves proximal to distal growth and secondary and

tertiary branching of the prostate ducts (38). To determine whether stromal ER␣ was involved in this prostate branching morphogenesis, mouse prostates were collected and microdissected. After microdissection, prostate tips, branches, and branch points were counted using a dissection microscope. The smER␣KO mice prostates do not show a significantly reduced number of prostate tips in the VP (Fig. 2, C and C-1). However, in agreement with previous reports from our lab, fER␣KO mice do show a decrease in the number of VP tips. Thus, we can conclude that fER␣ plays a role in branching morphogenesis but not in the prostate infolding or collagen deposition. ER␣ is able to regulate growth and collagen through IGF/EGF and matrix metalloproteinase 9 (MMP9) pathways To determine the mechanism through which ER␣ exerted its influence, rat PS-1 cells were used as an in vitro model. PS-1 cells have been characterized previously as being a prostate smooth muscle cell type (40). First, cells that expressed ER␣ were created and verified using immunofluorescent staining and Western blotting (Fig. 3, A and C). When ER␣ was introduced into PS-1 cells, these cells grew significantly faster than vector-infected cells (Supplemental Fig. 3). Working under the assumption that ER␣ was influencing epithelial cells in a paracrine


FIG. 2. Phenotypic changes of smER␣KO mice. A, Histological comparison of smER␣KO, fER␣KO, and Wt mouse prostates of 12 wk. Histological changes were present between the groups, hemotoxylin and eosin-stained smER␣KO, fER␣KO, and Wt mouse prostates at 12 wk of age were compared. Proximal (Prox) and distal (Dist) areas of the prostates were compared. The proximal area of the smER␣KO VP has a reduction of infolding as compared with the VPs of Wt and fER␣KO (arrows signify areas of magnification). A-1, Quantification data reveals a significant decrease in smER␣KO mouse VP infolding (P ⬍ 0.01). No significant change was seen in fER␣KO mice. B and B-1, Collagen deposition was measured using picrosirius red staining. smER␣KO mice show significantly less collagen fibrils than Wt mice or fER␣KO mice (P ⬍ 0.05). C and C-1, VPs were digested using collagenase, and the tips were counted and quantified. smER␣KO mice show no difference in the amount of tips compared with Wt, yet fER␣KO mice do (P ⬍ 0.05) (n ⫽ 3).


43

mechanism, CM from PS-1 cells with or without ER␣ was collected. This CM was used to treat BPH-1 epithelial cells and the growth effect observed (Fig. 3B). BPH-1 cells exposed to CM from PS-1 ER␣-infected cells grew significantly better than cells treated with PS-1 vector-infected CM. Several growth factors that are involved in prostate development were assayed, and Q-PCR results showed a significant increase of both EGF and IGF-I in ER␣⫹ cells (Fig. 3D). This increase was not due to differences in cell passage number between parental PS-1 cells and infected PS-1 cells (Supplemental Fig. 3). Previous studies have shown that E2 can up-regulate IGF-I in smooth muscle cells and that this upregulation can increase these cells’ proliferation (41). In addition to IGF, we have also observed that ER␣ can regulate EGF in other cell lines (Yeh, S., unpublished data). EGF has previously been linked to prostate cell differentiation through an androgen receptor (AR)-dependent axis (42). To determine whether IGF-I and EGF are direct target genes of ER␣, we performed a time course with 0, 6, and 24 h of 10 nM E2 treatments in PS-1 ER␣⫹ cells. At 6 h, no difference in EGF/IGF-I signal is seen, indicating that these genes are most likely not direct targets of ER␣. At 24 h, the levels of these two genes are significantly changed between vector and ER␣⫹ PS-1 cells, (Fig. 3E), further suggesting an indirect regulation. ER␣ has been linked to IGF-I expression in smooth muscle in a previous report (41) that showed decreased IGF-I function in smooth muscle cells isolated from Balb/C Wt mice, thus it remains unclear whether this regulation is direct. To investigate further, we performed promoter analysis of IGF-I using Dragon ERE software version 3 (http://datam.i2r.a-star.edu. sg/ereV3/index.html). This analysis did not identify any perfect ERE within 6 kb upstream of the IGF-I promoter, although one noncanonical site was

44

Vitkus et al.



FIG. 3. Validation of ER␣ roles and mechanism in rat smooth muscle PS-1 cells. A, PS-1 cells containing ER␣ were created and verified using immunofluorescence and Q-PCR. B, BPH-1 epithelial cells were treated with CM collected from PS-1⫹/⫺ ER␣ cells. When ER␣ is not present, the CM failed to stimulate prostate epithelial cell growth. C, MMP9 was increased in smooth muscle PS-1 cells lacking ER␣. Total MMP9 protein was detected using Western blotting and rabbit polyclonal MMP9 antibody (Abcam, Cambridge, MA). D, Q-PCR analysis of several growth factors. mRNA from PS-1 cells with or without ER␣ was collected 24 hours after treatment with 10nM E2. cDNA was obtained via reverse transcription from mRNA and qPCR analysis was performed using a panel of growth factors. E, PS-1 cells with or without ER␣ were treated with 10 nM E2, and RNAs from cell lysates were collected at various time points. Results are representative of three independent experiments. DAPI, 4⬘,6-Diamidino-2phenylindole; GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (*, P ⬍ 0.05; **, P ⬍ 0.01).

flagged (data not shown), which may account for the slight, but insignificant, increase in IGF-I expression seen at 6 h. Together, these data suggest that IGF-I and EGF are regulated in a secondary indirect fashion. It is important to note that in addition to IGF/EGF regulation, (C-C motif) ligand 2 (CCL2) is found to be down-regulated in ER␣⫹ PS-1 cells, whereas bone morphogenic protein 4

(BMP4) and AR show no significant change (Fig. 3D). Because a literature search provided no direct link between ER␣ and MMP9, it is possible that this regulation is through CCL2, which has been shown to up-regulate MMP9 in prostate cells (43). AR levels remain unchanged, so we can conclude that the growth phenotype observed was not caused via increased AR signaling. Sim-


ilarly, ER␤ protein amounts were not dramatically changed between vector and ER␣-infected PS-1 cells (Supplemental Fig. 3). Our previous reports with fER␣KO mice showed that BMP4 was a target of ER␣ responsible for branching morphogenesis (22). Here, we show that ER␣ is unable to change the levels of BMP4 in smooth muscle cells. These data are consistent with our mouse findings, where branching morphogenesis was unchanged in smER␣KO mice. MMP9 is a matrix metalloproteinase known to degrade components of the basement membrane (44, 45). To see whether MMP9 is involved in the ER␣-mediated collagen decrease, protein extracts from PS-1 cells with or without ER␣ were tested for MMP9 protein levels via Western blotting. Results show that ER␣ is able to inhibit MMP9 at the protein level (Fig. 3C). It is possible that through a pathway involving EGF/IGF, ER␣ is able to maintain normal prostate growth and also to maintain the basement membrane


45

integrity through down-regulation of MMP9, possibly through a CCL2-related mechanism. Mice lacking ER␣ in both smooth muscle and fibroblasts develop an aggregate prostate phenotype Mice lacking ER␣ in both smooth muscle and the fibroblasts were created (dER␣KO) by mating Tgln cre with FSP cre and the resultant offspring with ER␣ F/F mice to obtain offspring mice displaying all three genotypes (Supplemental Fig. 4). In general, the dER␣KO mice displayed a compound phenotype of both the fER␣KO and smER␣KO mice. dER␣KO mice VPs displayed a decrease in the number of ventral tips, which is found to be significant when compared with Wt and is similar to the decrease seen in fER␣KO mice (Fig. 4, A and C). When gross histology was looked at to determine infolding amounts, the same phenotype as the smER␣KO mice was

FIG. 4. dER␣KO mouse prostates display an aggregate phenotype of fER␣KO and smER␣KO mice. A and C, dER␣KO mouse VPs display reduced branching morphogenesis similar to fER␣KO mice when compared with prostates of age-matched Wt littermates. (P ⬍ 0.01, n ⫽ 4). B and D, dER␣KO mouse prostates have significantly decreased glandular infolding compared with that of age-matched Wt littermates (P ⬍ 0.01). E and F, Collagen deposition is also decreased in prostates of dER␣KO mice, similar to smER␣KO mice. The decrease was found to be significant compared with Wt mice prostates (P ⬍ 0.05) (n ⫽ 6). Prox, Proximal; Dist, distal.

46

Vitkus et al.


observed in the dER␣KO mice. dER␣KO mice show a significant decrease in the proximal epithelial infolding compared with Wt mice (Fig. 4, B and D). Finally, collagen deposition was also significantly decreased in dER␣KO mice compared with Wt mice, and this was consistent with smER␣KO mice (Fig. 4, E and F).

Discussion The present study shows that ER␣ acts through both smooth muscle and fibroblast cells to influence the prostate epithelial cells. It has been shown that smooth muscle cells can have an influence on epithelial cells in the context of development (46). However, it remains unclear in what manner ER␣ may affect this cross talk. Studies using ubiquitous ER␣KO mice have shown that ER␣ plays a role in the prostate organogenesis (27), and here, we show how, specifically, two stromal cell types differentially influence the prostate organ homeostasis. Additionally, epithelial ER␣KO mice via prostate epithelial-specific probasin-cre show no branching morphogenesis abnormalities (47). Specifically, our smER␣KO model demonstrates that ER␣ plays a supporting role in the maintenance of the basement membrane, of glandular infolding, and in gross prostate morphology in a cell type-specific manner. Originally, it was expected that ER␣KO in both smooth muscle and fibroblast cells would show an even


more severe decrease in the VP branching morphogenesis than was seen in the fER␣KO alone. However, instead of showing a more severe singular phenotype, dER␣KO mice showed an overall more severe defect of prostate development. This dual ER␣ role led to prostates of dER␣KO mice demonstrating a combination of the phenotypes seen from fER␣KO and smER␣KO mice. Smooth muscle cells are two to four layers thick around the proximal area and thinner toward the distal area with a thin layer of periductal fibroblasts between them and the epithelial cells (33). Our data suggest that ER␣ in the smooth muscle cells plays an important role in epithelial cell differentiation in a spatial-dependent manner. In the proximal area, where smooth muscle cells are prevalent, ER␣ is able to easily exert its influence to assist in normal epithelial cell differentiation, but in the distal area, smooth muscle cells are not as abundant, so ER␣ may not play as important of a role. From this result, we hypothesize that potential growth factors from the fibroblasts, and not smooth muscle cells, which are not regulated by ER␣, could help to maintain the normal epithelial differentiation pattern. In normal prostate development, ER␣ and smooth muscle IGF/EGF may cooperate to influence the glandular development (Fig. 5). Our group has previously shown that fER␣KO mice show a decrease in branching morphogenesis as adults as well as a host of other urogenital phenotypes. However, it would seem

FIG. 5. ER␣ in different stromal cell types is able to exert differential effects on prostate epithelial cells. The figure depicts that ER␣ in smooth muscle is able to up-regulate IGF-I and EGF levels to influence epithelial cell growth, whereas down-regulation of MMP9 maintains basement membrane integrity. Stromal fER␣ through inhibition of BMP4 is able to decrease prostate branch morphogenesis.


that although fER␣ plays an important role in branching morphogenesis, smER␣ is more important for maintenance of the glandular structure and basement membrane thickness. Of note is that estrogenization of neonatal male mice results in an increase of periductal fibroblasts with the end result being a decrease in branching morphogenesis. The purported mechanism for this phenotype is through constraint of epithelial cells (7). In our fER␣KO mice, a different mechanism through BMP4 regulation and stromal apoptosis is able to achieve a similar phenotype (22). Interestingly, when we assayed this gene, it was unchanged in the smER␣KO mice (Supplemental Fig. 5). If a gene is up-regulated in fER␣KO mice and downregulated or unchanged in smER␣KO mice, the net sum of the expression would display in the dER␣KO mice. This effect may mask the importance of ER␣ in prostate gene expression. For this reason, it is important to separate out the fibroblast and smooth muscle cell ER␣ function. By separating out the roles of the smooth muscle and fER␣, we have exhibited the power of the cre/lox system. Previous research into the role of ER␣ in prostate development has used tissue recombination using tissues derived from ER␣ total KO or Wt mice. Our findings demonstrate some similar phenotypes, because results from these studies have shown that ER␣ in the stromal compartment is important for prostate development (48). However, these models use systems where ER is removed from every cell type in the stroma, including endothelial cells, neuroendocrine cells, and, importantly, both smooth muscle and fibroblast cells, which are important for two different sets of phenotypes. Recent evidence has implicated the endothelial cells as an important factor for promoting prostate tumor metastasis (49). Using an endothelial cell-specific promoter, we could create a mouse model where only endothelial cell ER␣ is lost to discover the potential role of ER␣ in modulating endothelial cellmediated prostate cancer invasion. A major drawback of the cre/lox mouse model is that complete removal of the target gene is often impossible. In our model, we see that there is 50% cre activity in the VP (Supplemental Fig. 6). This can pose a problem when interpreting data, because a partial KO may have a different phenotype than that in either a complete KO or Wt mouse. One good example of this is the TGF␤ receptor II (TGF␤R2). KO of TGF␤R2, using loxP sites at introns 1 and 2 of Tgf␤r2 and cre recombinase driven by the FSP promoter, resulted in prostatic intraepithelial neoplasia (PIN) lesions in the prostate. From this, Bhowmick et al. (50) concluded that the loss of TGF␤ signaling via KO of TGF␤R2 resulted in PIN lesions. In a tissue recombination model, Franco et al. (51) created a mixture of 50% normal prostate stromal


47

cells and 50% normal cells where the TGF␤R2 had been knocked out. When these cells were combined with normal epithelial cells, the resultant tissue recombinant was able to create adenocarcinoma (51). This finding was later expanded and explained in a similar system when TGF␤R2 was completely knocked out, left intact, or modulated to 50% KO. In the intact group, benign prostate growth was increased, and in the total KO group, PIN lesions were observed (similar to FSP-TGF␤R2 mice). In the 50% KO group, adenocarcinoma was observed (52). These data clearly present one of the shortcomings of transgenic mouse models, incomplete KO may lead to different phenotypes. The role of ER␣ and ER␤ in the initiation and progression of prostate cancer has been studied (reviewed in Refs. 28, 53), but only recently has evidence begun to come to light about whether stromal or epithelial ER␣ is important in this transition (54). Epithelial ER␣ has been shown to have no effect on prostate development at 12 wk but plays an important role in the development of squamous metaplasia (47). KO of ER␣ in the stromal could lead to a decrease in cancer growth/initiation. Perhaps more interesting would be the effect of smER␣KO on prostate metastasis due to ER␣’s role in basement membrane integrity. Stromal ER␣ is already implicated in BPH initiation (10). This is especially poignant, because BPH is characterized by an increase in smooth muscle cells (55). The IGF pathway has been studied extensively in normal and malignant prostate. IGF is important for normal cell proliferation and when up-regulated can lead to malignant transformation (56). Here, we show that ER␣ is able to regulate this protein’s expression in a rat smooth muscle cell line and also at a gene level in a transgenic mouse model. Disruption of the IGF pathway during development could help to explain some of the phenotypes seen in our mouse model, such as the decreased infolding and decreased branching morphogenesis (27). EGF has been shown to act as a stimulator of cell growth through the MAPK/ERK pathway (57). EGF has been shown to mediate the ER␣ pathway in ovary (58), but it is unclear whether this interaction is found in the prostate. Further studies will reveal the intricate mechanisms behind EGF/ ER␣-mediated cell growth. The dER␣KO mice show an aggregate of the phenotypes observed in the smooth muscle and fibroblast cell types, indicating a distinct role for ER␣ in two separate stromal cell types. Our in vitro and in vivo data confirm the ability of smooth muscle ER␣ to modulate IGF-I and EGF expression, which has been shown to modulate stromal cell proliferation, potentially leading to the phenotype changes observed in this article.

48

Vitkus et al.



Acknowledgments We thank Karen Wolf for her help in manuscript preparation. Address all correspondence and requests for reprints to: Shuyuan Yeh, University of Rochester Medical Center, 601 Elmwood Drive, Rochester, New York 14642. E-mail: [email protected]. This work was supported by the National Institutes of Health Grant CA137474. Disclosure Summary: The authors have nothing to disclose.

References 1. Kuiper GG, Carlsson B, Grandien K, Enmark E, Häggblad J, Nilsson S, Gustafsson JA 1997 Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors ␣ and ␤. Endocrinology 138:863– 870 2. Ellem SJ, Risbridger GP 2009 The dual, opposing roles of estrogen in the prostate. Ann NY Acad Sci 1155:174 –186 3. Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA 1996 Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 93:5925–5930 4. Cooke PS, Young P, Hess RA, Cunha GR 1991 Estrogen receptor expression in developing epididymis, efferent ductules, and other male reproductive organs. Endocrinology 128:2874 –2879 5. Prins GS, Birch L, Couse JF, Choi I, Katzenellenbogen B, Korach KS 2001 Estrogen imprinting of the developing prostate gland is mediated through stromal estrogen receptor ␣: studies with ␣ERKO and ␤ERKO mice. Cancer Res 61:6089 – 6097 6. Chen M, Hsu I, Wolfe A, Radovick S, Huang K, Yu S, Chang C, Messing EM, Yeh S 2009 Defects of prostate development and reproductive system in the estrogen receptor-␣ null male mice. Endocrinology 150:251–259 7. Chang WY, Wilson MJ, Birch L, Prins GS 1999 Neonatal estrogen stimulates proliferation of periductal fibroblasts and alters the extracellular matrix composition in the rat prostate. Endocrinology 140:405– 415 8. Folkerd EJ, Dowsett M 2010 Influence of sex hormones on cancer progression. J Clin Oncol 28:4038 – 4044 9. Celhay O, Yacoub M, Irani J, Dore B, Cussenot O, Fromont G 2010 Expression of estrogen related proteins in hormone refractory prostate cancer: association with tumor progression. J Urol 184:2172– 2178 10. Park II, Zhang Q, Liu V, Kozlowski JM, Zhang J, Lee C 2009 17␤-estradiol at low concentrations acts through distinct pathways in normal versus benign prostatic hyperplasia-derived prostate stromal cells. Endocrinology 150:4594 – 4605 11. Yang R, Ma YX, Chen LF, Zhou Y, Yang ZP, Zhu Y, Du XL, Shi JD, Ma HS, Zhang J 2010 Antagonism of estrogen-mediated cell proliferation by raloxifene in prevention of ageing-related prostatic hyperplasia. Asian J Androl 12:735–743 12. Farnsworth WE 1999 Estrogen in the etiopathogenesis of BPH. Prostate 41:263–274 13. vom Saal FS, Timms BG, Montano MM, Palanza P, Thayer KA, Nagel SC, Dhar MD, Ganjam VK, Parmigiani S, Welshons WV 1997 Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses. Proc Natl Acad Sci USA 94:2056 –2061 14. Cunha GR, Hayward SW, Wang YZ 2002 Role of stroma in carcinogenesis of the prostate. Differentiation 70:473– 485 15. Niu YN, Xia SJ 2009 Stroma-epithelium crosstalk in prostate cancer. Asian J Androl 11:28 –35 16. Basanta D, Strand DW, Lukner RB, Franco OE, Cliffel DE, Ayala

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32. 33. 34.

35.

GE, Hayward SW, Anderson AR 2009 The role of transforming growth factor-␤-mediated tumor-stroma interactions in prostate cancer progression: an integrative approach. Cancer Res 69:7111– 7120 Zhao FJ, Han BM, Yu SQ, Xia SJ 2009 Tumor formation of prostate cancer cells influenced by stromal cells from the transitional or peripheral zones of the normal prostate. Asian J Androl 11:176 – 182 Ogawa S, Lubahn DB, Korach KS, Pfaff DW 1996 Aggressive behaviors of transgenic estrogen-receptor knockout male mice. Ann NY Acad Sci 794:384 –385 Kyprianou N, Isaacs JT 1988 Activation of programmed cell death in the rat ventral prostate after castration. Endocrinology 122:552– 562 Justulin Jr LA, Ureshino RP, Zanoni M, Felisbino SL 2006 Differential proliferative response of the ventral prostate and seminal vesicle to testosterone replacement. Cell Biol Int 30:354 –364 Lesser B, Bruchovsky N 1974 The effects of 5␣-dihydrotestosterone on the kinetics of cell proliferation in rat prostate. Biochem J 142:483– 489 Chen M, Yeh CR, Shyr CR, Lin HH, Da J, Yeh S 2012 Reduced prostate branching morphogenesis in stromal fibroblast, but not in epithelial, estrogen receptor ␣ knockout mice. Asian J Androl 14: 546 –555 Mehasseb MK, Bell SC, Habiba MA 2009 The effects of tamoxifen and estradiol on myometrial differentiation and organization during early uterine development in the CD1 mouse. Reproduction 138:341–350 Alda JO, Valero MS, Pereboom D, Gros P, Garay RP 2009 Endothelium-independent vasorelaxation by the selective ␣ estrogen receptor agonist propyl pyrazole triol in rat aortic smooth muscle. J Pharm Pharmacol 61:641– 646 Smiley DA, Khalil RA 2009 Estrogenic compounds, estrogen receptors and vascular cell signaling in the aging blood vessels. Curr Med Chem 16:1863–1887 Odenlund M, Ekblad E, Nilsson BO 2008 Stimulation of oestrogen receptor-expressing endothelial cells with oestrogen reduces proliferation of cocultured vascular smooth muscle cells. Clin Exp Pharmacol Physiol 35:245–248 Chen M, Wolfe A, Wang X, Chang C, Yeh S, Radovick S 2009 Generation and characterization of a complete null estrogen receptor ␣ mouse using Cre/LoxP technology. Mol Cell Biochem 321: 145–153 Prins GS, Korach KS 2008 The role of estrogens and estrogen receptors in normal prostate growth and disease. Steroids 73:233– 244 Holtwick R, Gotthardt M, Skryabin B, Steinmetz M, Potthast R, Zetsche B, Hammer RE, Herz J, Kuhn M 2002 Smooth muscleselective deletion of guanylyl cyclase-A prevents the acute but not chronic effects of ANP on blood pressure. Proc Natl Acad Sci USA 99:7142–7147 Yu S, Zhang C, Lin CC, Niu Y, Lai KP, Chang HC, Yeh SD, Chang C, Yeh S 2011 Altered prostate epithelial development and IGF-1 signal in mice lacking the androgen receptor in stromal smooth muscle cells. Prostate 71:517–524 Cunha GR, Hayward SW, Wang YZ, Ricke WAS 2003 Role of stromal microenvironment in carcinogenesis of the prostate. Int J Cancer 107:1–10 Liu AY, True LD 2002 Characterization of prostate cell types by CD cell surface molecules. Am J Pathol 160:37– 43 Flickinger CJ 1972 The fine structure of the interstitial tissue of the rat prostate. Am J Anat 134:107–125 Rouleau M, Léger J, Tenniswood M 1990 Ductal heterogeneity of cytokeratins, gene expression, and cell death in the rat ventral prostate. Mol Endocrinol 4:2003–2013 Kassen A, Sutkowski DM, Ahn H, Sensibar JA, Kozlowski JM, Lee


36.

37. 38.

39.

40.

41.

42.

43.

44.

45. 46.

47.


C 1996 Stromal cells of the human prostate: initial isolation and characterization. Prostate 28:89 –97 Solway J, Seltzer J, Samaha FF, Kim S, Alger LE, Niu Q, Morrisey EE, Ip HS, Parmacek MS 1995 Structure and expression of a smooth muscle cell-specific gene, SM22␣. J Biol Chem 270:13460 – 13469 Singh J, Zhu Q, Handelsman DJ 1999 Stereological evaluation of mouse prostate development. J Androl 20:251–258 Marker PC, Donjacour AA, Dahiya R, Cunha GR 2003 Hormonal, cellular, and molecular control of prostatic development. Dev Biol 253:165–174 Stanley JR, Woodley DT, Katz SI, Martin GR 1982 Structure and function of basement membrane. J Invest Dermatol 79(Suppl 1): 69s–72s Gerdes MJ, Dang TD, Lu B, Larsen M, McBride L, Rowley DR 1996 Androgen-regulated proliferation and gene transcription in a prostate smooth muscle cell line (PS-1). Endocrinology 137:864 – 872 Zhou D, Li S, Wang X, Cheng B, Ding X 2011 Estrogen receptor ␣ is essential for the proliferation of prostatic smooth muscle cells stimulated by 17␤-estradiol and insulin-like growth factor 1. Cell Biochem Funct 29:120 –125 Hitosugi T, Sasaki K, Sato M, Suzuki Y, Umezawa Y 2007 Epidermal growth factor directs sex-specific steroid signaling through Src activation. J Biol Chem 282:10697–10706 Herman JG, Stadelman HL, Roselli CE 2009 Curcumin blocks CCL2-induced adhesion, motility and invasion, in part, through down-regulation of CCL2 expression and proteolytic activity. Int J Oncol 34:1319 –1327 Zeng ZS, Cohen AM, Guillem JG 1999 Loss of basement membrane type IV collagen is associated with increased expression of metalloproteinases 2 and 9 (MMP-2 and MMP-9) during human colorectal tumorigenesis. Carcinogenesis 20:749 –755 Sternlicht MD, Werb Z 2001 How matrix metalloproteinases regulate cell behavior. Annu Rev Cell Dev Biol 17:463–516 Cunha GR, Hayward SW, Dahiya R, Foster BA 1996 Smooth muscle-epithelial interactions in normal and neoplastic prostatic development. Acta Anat (Basel) 155:63–72 Chen M, Yeh CR, Chang HC, Vitkus S, Wen XQ, Bhowmick NA, Wolfe A, Yeh S 2012 Loss of epithelial oestrogen receptor ␣ inhibits

48. 49.

50.

51.

52.

53.

54.

55.

56. 57.

58.

49

oestrogen-stimulated prostate proliferation and squamous metaplasia via in vivo tissue selective knockout models. J Pathol 226: 17–27 Cunha GR, Cooke PS, Kurita T 2004 Role of stromal-epithelial interactions in hormonal responses. Arch Histol Cytol 67:417– 434 Loberg RD, Day LL, Harwood J, Ying C, St John LN, Giles R, Neeley CK, Pienta KJ 2006 CCL2 is a potent regulator of prostate cancer cell migration and proliferation. Neoplasia 8:578 –586 Bhowmick NA, Chytil A, Plieth D, Gorska AE, Dumont N, Shappell S, Washington MK, Neilson EG, Moses HL 2004 TGF-␤ signaling in fibroblasts modulates the oncogenic potential of adjacent epithelia. Science 303:848 – 851 Franco OE, Jiang M, Strand DW, Peacock J, Fernandez S, Jackson 2nd RS, Revelo MP, Bhowmick NA, Hayward SW 2011 Altered TGF-␤ signaling in a subpopulation of human stromal cells promotes prostatic carcinogenesis. Cancer Res 71:1272–1281 Kiskowski MA, Jackson 2nd RS, Banerjee J, Li X, Kang M, Iturregui JM, Franco OE, Hayward SW, Bhowmick NA 2011 Role for stromal heterogeneity in prostate tumorigenesis. Cancer Res 71: 3459 –3470 Bonkhoff H, Berges R 2009 The evolving role of oestrogens and their receptors in the development and progression of prostate cancer. Eur Urol 55:533–542 Hu WY, Shi GB, Lam HM, Hu DP, Ho SM, Madueke IC, KajdacsyBalla A, Prins GS 2011 Estrogen-initiated transformation of prostate epithelium derived from normal human prostate stem-progenitor cells. Endocrinology 152:2150 –2163 Marks LS, Treiger B, Dorey FJ, Fu YS, deKernion JB 1994 Morphometry of the prostate: I. Distribution of tissue components in hyperplastic glands. Urology 44:486 – 492 Baserga R 1995 The insulin-like growth factor I receptor: a key to tumor growth? Cancer Res 55:249 –252 Fallon JH, Seroogy KB, Loughlin SE, Morrison RS, Bradshaw RA, Knaver DJ, Cunningham DD 1984 Epidermal growth factor immunoreactive material in the central nervous system: location and development. Science 224:1107–1109 Ankrapp DP, Bennett JM, Haslam SZ 1998 Role of epidermal growth factor in the acquisition of ovarian steroid hormone responsiveness in the normal mouse mammary gland. J Cell Physiol 174: 251–260

Submit your best science for presentation in an oral or poster session at ENDO 2013, June 15 - 18, 2013, San Francisco, California Abstract Submission Deadline: January 30, 2013 12:00PM (noon) ET www.endo-society.org/endo2013