European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Increased expression of fibroblast growth factor receptor 1 in endometriosis and its correlation with endometriosis-related dysmenorrhea and recurrence Linjie Zhao a,b, Huiliang Yang a,b, Yu Xuan a,b, Zhongyue Luo c, Qiao Lin c, Jitong Zhao a,b, Ning Ren a,b, Shengtao Zhou a,b,*, Xia Zhao a,b a
Department of Gynecology and Obstetrics, Key Laboratory of Obstetrics & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second Hospital, Chengdu 610041, PR China b The State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Chengdu 610041, PR China c College of Life Science, Sichuan University, Chengdu 610041, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 June 2014 Received in revised form 23 October 2014 Accepted 11 November 2014
Objective(s): This study aims to identify a critical molecule that potentially participates in endometriosis pathogenesis and characterize its correlation with dysmenorrhea and recurrence. Study design: We utilized a bioinformatics-based strategy to screen for candidate genes and fibroblast growth factor receptor 1(FGFR1) was chosen for further validation. FGFR1 expression was examined in specimens of ectopic and eutopic endometrium obtained from 48 patients with endometriosis and specimens of eutopic endometrium from 26 healthy control subjects using immunohistochemistry and Western blotting. In addition, FGFR shRNA treatment was applied in a nude mice endometriosis model to examine the functional role of FGFR1 in endometriosis formation in vivo. Results: FGFR1 was found commonly overexpressed in ectopic endometrium of endometriosis compared with either its eutopic counterpart or endometrium from normal patients (P < 0.05). FGFR shRNA treatment impaired endometriosis formation and alleviated endometriosis-related symptoms in vivo. FGFR1 expression in ectopic endometrium was correlated with dysmenorrhea severity (P < 0.05) and recurrence in endometriosis patients (P < 0.05). Conclusion(s): FGFR1 might be involved in endometriosis development, which could possibly serve as a novel therapeutic target and prognostic marker for this disease. ß 2014 Published by Elsevier Ireland Ltd.
Keywords: Endometriosis FGFR1 Dysmenorrhea Recurrence
Introduction Endometriosis is a common estrogen-dependent gynecological disorder in females of reproductive age and is characterized by the presence and growth of endometrial tissue outside the uterus [1]. The development of ectopic endometrial tissues suggests intrinsic cellular mechanisms leading to invasion, unrestrained growth, neoangiogenesis, and distant spreading of endometriotic cells [2]. Despite its high prevalence, the etiology and pathogenesis of endometriosis still remain largely unelucidated. Moreover, although hormonal therapy offers an alternative for the treatment of endometriosis, its curative efficacy for endometriosis is often
* Corresponding author at: Department of Gynecology and Obstetrics, West China Second Hospital, Sichuan University, Chengdu, 610041, PR China. Tel.: +86 28 13551070137.. E-mail address: [email protected] (S. Zhou). http://dx.doi.org/10.1016/j.ejogrb.2014.11.013 0301-2115/ß 2014 Published by Elsevier Ireland Ltd.
insufficient. Therefore, the development of novel treatment strategies for endometriosis based upon new potential therapeutic targets is imperative. Previous observations supported the notion that endometriosis might be a ‘‘metastasis’’-related disease, which offers endometriosis some similarities to cancer diseases although metastasis does not really exist in endometriosis. However, the underlying mechanisms are still unclear [3,4]. Fibroblast growth factor receptor (FGFR) family is part of a group of growth factor receptor tyrosine kinases (RTK) able to induce several cellular processes including cell proliferation, angiogenesis, inhibition of apoptosis, and cell migration [5]. FGFR1, as a member of fibroblast growth factor receptors, is known to drive an epithelial-to-mesenchymal transition (EMT) of primitive streak-localized epiblast cells into mesoderm cells. During EMT, epithelial cells lose their polarity, augment their motility, and begin to express mesenchymal markers, such as vimentin, becoming ‘‘mesenchymal-like’’. This process has also been closely linked to cancer progression and
118
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
metastasis. Previously, aberrant FGFR1 expression has been implicated in the progression of a variety of tumor cell types including prostate cancer, colorectal cancer and non-small lung cancer [5]. However, the contribution of FGFR1 to the pathogenesis of benign disorders like endometriosis is still unclear. Thus, our study attempted to identify a critical molecule that might participate in endometriosis development. Materials and methods Bioinformatics analysis Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI) was searched to identify candidate genes responsible for endometriosis development. Three GEO datasets analyzing endometriosis were identified, including GSE5108, GSE6364, and GSE7305 [6–8]. These three datasets were compared for overlapped genes differentiallyexpressed between endometriosis and control. Two criteria were applied: (a) significantly differentially-regulated (P < 0.05) between endometriosis and control identified by GEO, (b) significantly differentially-expressed by at least 2-fold (P < 0.05). Since endometriosis might be a ‘‘metastasis’’-related disease similar to cancer although metastasis does not really exist, we further searched eligible genes against KEGG pathway database, Gene Ontology and metastasis-related published studies to identify genes most closely associated with metastasis. These ‘‘metastasis’’ genes were subsequently analyzed in STRING software, an open source web-based tool with established and predicted protein interactions, and visualized protein-protein interaction networks [9]. Those key nodes with over two interactive nodes demonstrated by STRING pathway analysis were selected as potential candidate genes responsible for the pathogenesis of endometriosis. Patients and tissue samples Eutopic endometrium used for Western blot analysis was obtained from 26 patients undergoing hysterectomy for benign indications other than endometriosis (like leiomyoma) as control, and ectopic endometrium was obtained from endometriosis patients at the Gynecological Department of West China Second Hospital of Sichuan University (Chengdu, China) from 2009 to 2014. None of the patients had received any preoperative hormonal therapy prior to surgery. All these samples were obtained by experienced gynecologists and gynecological surgeons and examined by experienced pathologists who confirmed the diagnosis of disease samples. Paraffin-embedded eutopic endometrial and paired ectopic endometrial specimens were obtained from 48 patients who underwent surgical resections in the same hospital from 2009 to 2014. They were further subjected to immunohistochemical analysis of identified proteins for validation and subsequent follow-up were carried out for these 48 patients. This study was approved by the Institutional Ethics Committee of Sichuan University. Informed consents were obtained from all patients prior to analysis. Laser capture microdissection (LCM) and Western blotting All lesions were microdissected from 12-mm-thick frozen sections using LCM system (Leica, Germany). Normal endometrial tissues were dissected to obtain negative control proteins. Western blotting analysis was conducted as previously described [10]. The primary antibody was rabbit anti-FGFR1 (Santa Cruz Biotechnology). b-Actin was used as an internal control.
(Santa Cruz Biotechnology) and rabbit anti-Ki-67 (Santa Cruz Biotechnology). Slides were evaluated by two independent pathologists in a double-blinded manner. Any discrepancy between the two evaluators was resolved by reevaluation and careful discussion until agreement was reached. Animal model of endometriosis The guidelines for animal care were approved by the Institutional Animal Care and Use Committee of Sichuan University (Chengdu, Sichuan, People’s Republic of China). Endometriosis lesions were acquired from premenopausal women with endometriosis at West China Second Hospital, Sichuan University. In vivo endometriosis nude mice model was surgically induced as described previously [11]. In particular, each mouse received an intraperitoneal injection of PBS containing a suspension of five human endometriotic tissue fragments with a size of about 10 mm3 per mouse into the ventral midline and subcutaneous injection of 0.5 mg of 17b-estradiol was performed on days 1 and 2 to facilitate the implantation of endometriotic lesions. Mice were next assigned randomly to one of the following groups (5 per group): (a) PBS, 100 ml of PBS; (b) Lipo, Lipofectamine 2000 (Invitrogen) at 62.5 mg/100 ml of PBS; (c) shFGFR1, FGFR1 shRNA at 25 mg/100 ml of PBS. Intraperitoneal treatment was initiated 5 days after inoculation. Mice received therapy, three times a week and were sacrificed at 21 days postinoculation. Intraperitoneal endometrial nodules were resected and measured immediately to assess the treatment efficacy as previously described [10]. Pain assessment and follow-up of endometriosis patients Pain assessment of endometriosis patients was conducted as described elsewhere [12]. The 48 patients with endometriosis were further followed up for recurrence for 30 months after surgery. The recurrence of endometriosis was defined as: (a) the presence of ovarian cysts of 3 cm in diameter, along with characteristic echoes as detected by transvaginal ultrasonography for two consecutive menstrual cycles, coupled with or without the recurrence of dysmenorrhea or pelvic pain requiring medical intervention, or as (b) the presence of de novo endometriosis as confirmed by histology following a second surgery. We also included endometrial tissue samples from 26 women with other benign gynecological diseases as controls and none of them had endometriosis or adenomyosis. Statistics Data are presented as mean SD of three independent experiments unless otherwise indicated. GraphPad Prism (GraphPad Software Inc., CA, USA) was used for data analysis. The comparison of distributions of continuous variables was made using the Wilcoxon test or Kruskal–Wallis test. Jonckheere–Terpstra trend test was utilized to test for trend of FGFR1 immunoreactivity in women reporting more severe dysmenorrhea as described previously [13]. The correlation between FGFR1 staining scores and dysmenorrhea scores was analyzed using Pearson’s x2 test. Comparisons between two groups were performed with the Student’s t test, and differences among multiple groups were evaluated by one-way ANOVA analysis. Differences were considered statistically significant at P < 0.05. Results
Immunohistochemistry
Bioinformatics identification of FGFR1 as a candidate gene for endometriosis development
Immunohistochemistry (IHC) was performed as described previously [10]. The primary antibodies included rabbit anti-FGFR1
A total of 493 genes significantly differentially-regulated by at least 2-fold (P < 0.05) and overlapped in three datasets were
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
identified (Fig. 1A and B). Among the 493 overlapped genes, 43 genes were further identified to be ‘‘metastasis-related’’. The detailed information of each identified gene was listed in Table 1 and this gene repertoire fell into distinct categories based upon their sub-cellular localizations (Fig. 1C) and biological functions (Fig. 1D). Subsequently, these genes were further subjected to
119
STRING pathway analysis and an interaction network was visualized (Fig. 1E). Based upon the above bioinformatics analysis, FGFR1 was selected for further functional study due to (a) the unclear role of FGFR1 in the pathogenesis of endometriosis; (b) its ‘‘plasma’’ membrane location, which facilitates it to sense extracellular
Fig. 1. Bioinformatics analysis of differentially expressed genes in ectopic and eutopic endometrium of human endometriosis. (A) Screening strategy for key molecules for endometriosis pathogenesis. (B) Venn diagram for overlapped genes of the three publicly available endometriosis datasets. (C) Gene expression cluster map generated by Cluster software. Expression of genes in the eutopic endometrium is constant at 0, whereas genes up-regulated in the ectopic endometrium are in red, and the downregulated proteins are in green. The intensity of the color green or red corresponds to the degree of alteration, respectively, according to the color strip at the bottom of the figure. (D) The identified genes were categorized into several protein groups according to their subcellular locations. (E) The identified genes were categorized into several protein groups according to their biological functions. (F) The protein-protein interaction network of identified genes analyzed by STRING software. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
120
Table 1 Overlapped metastasis-related genes in the three publicly available datasets. Gene no.
Accession no.
Protein name
Gene name
Average alteration in three datasets (fold)
Subcellular location
Function
1 2 3
P14651 Q12802 P63241
HOXB3 AKAP13 EIF5A
6.255 1.4 1.26
Nucleus Cytoplasm Cytoplasm
Angiogenesis Apoptosis Apoptosis
4 5 6 7
P40240 Q75L79 P43121 Q8IZT6
CD9 CLDN3 MCAM ASPM
1.41 6.87 1.68 4.655
Membrane Membrane Membrane Cytoplasm
Cell Cell Cell Cell
8 9
P48729 P23921
1.505 2.245
Cytoplasm Cytoplasm
Cell cycle Cell cycle
10 11 12 13 14 15 16 17 18
O14818 Q96GD4 Q9UNL4 P15923 P62826 P49815 P16949 Q8WVJ9 P25490
Cytoplasm Nucleus Nucleus Nucleus Nucleus Cytoplasm Cytoplasm Nucleus Nucleus
Cell Cell Cell Cell Cell Cell Cell Cell Cell
19
O75030
20 21
Q8NI77 O60271
22 23 24 25 26 27 28 29 30
P14923 P42345 P04626 P21246 P11362 P57735 P03372 Q9BXP5 P67809
31
Q9BY41
Homeobox protein Hox-B3 A-kinase anchor protein 13 Eukaryotic translation initiation factor 5A-1 CD9 antigen Claudin 3 Cell surface glycoprotein MUC18 Abnormal spindle-like microcephaly-associated protein Casein kinase I isoform alpha Ribonucleoside-diphosphate reductase large subunit Proteasome subunit alpha type-7 Aurora kinase B Inhibitor of growth protein 4 Transcription factor E2-alpha GTP-binding nuclear protein Ran Tuberin Stathmin Twist-related protein 2 Transcriptional repressor protein YY1 Microphthalmia-associated transcription factor Kinesin-like protein KIF18A C-Jun-amino-terminal kinase-interacting protein 4 Junction plakoglobin Serine/threonine-protein kinase mTOR Receptor tyrosine-protein kinase erbB-2 Pleiotrophin Fibroblast growth factor receptor 1 Ras-related protein Rab-25 Estrogen receptor Serrate RNA effector molecule homolog Nuclease-sensitive element-binding protein 1 Histone deacetylase 8
32
P24347
Stromelysin-3
MMP11
4.53
Secreted
33
P19235
Erythropoietin receptor
EPOR
1.54
Membrane
34 35 36 37
Q06455 B0YJ81 P00374 Q9HCC0
RUNX1T1 PTPLA DHFR MCCC2
38
Q16630
39
P62310
40
Q15599
41
Q92766
42
O76094
43
P05162
Protein CBFA2T1 3-hydroxyacyl-CoA dehydratase 1 Dihydrofolate reductase Methylcrotonoyl-CoA carboxylase beta chain, mitochondrial Cleavage and polyadenylation specificity factor subunit 6 U6 snRNA-associated Sm-like protein LSm3 Na(+)/H(+) exchange regulatory cofactor NHE-RF2 Ras-responsive element-binding protein 1 Signal recognition particle 72 kDa protein Galectin-2
signals and conduct information into individual cells; (c) its function in mediating cell migration under physiological conditions and tumor cell metastasis. Overexpression of FGFR1 in the ectopic endometrium of endometriosis To determine FGFR1 expression in human endometriotic tissues, we performed LCM to obtain endometrial tissues of paired eutopic and ectopic endometrium of endometriosis patients and
CSNK1A1 RRM1 PSMA7 AURKB ING4 TCF3 RAN TSC2 STMN1 TWIST2 YY1
1.24 3.24 1.13 1.94 1.64 1.26 2.03 3.05 1.16
adhesion adhesion adhesion cycle
cycle cycle cycle cycle cycle cycle differentiation differentiation differentiation
MITF
1.885
Nucleus
Cell differentiation
KIF18A SPAG9
3.205 1.395
Cytoplasm Cytoplasm
Cell migration Cell migration
JUP MTOR ERBB2 PTN FGFR1 RAB25 ESR1 SRRT YBX1
2.725 1.29 1.255 2.51 1.616667 4.37333 5.26 1.3 1.13
Membrane Cytoplasm Cytoplasm Secreted Membrane Membrane Nucleus Nucleus Cytoplasm
Cell Cell Cell Cell Cell Cell Cell Cell Cell
HDAC8
1.71
Nucleus
Nucleus Cytoplasm Cytoplasm Cytoplasm
Chromatin assembly or disassembly Collagen catabolic process Erythropoietin receptor activity Metabolism Metabolism Metabolism Metabolism
1.53 1.84 1.86 1.95333
migration proliferation proliferation proliferation proliferation proliferation proliferation proliferation proliferation
CPSF6
1.36333
Nucleus
mRNA processing
LSM3
1.32
Nucleus
mRNA processing
SIP1
1.38
Membrane
RREB1
1.295
Nucleus
Protein complex assembly Signal transduction
SRP72
1.32
Cytoplasm
Signal transduction
2.87667
Cytoplasm
Un known
LGALS2
control for further Western blotting analysis. Elevated FGFR1 expression was detected in ectopic endometrium of endometriosis patients compared with normal endometrium and eutopic endometrium of endometriosis (Fig. 2A and B). In addition, FGFR1 expression was validated in paraffin-embedded specimens of endometriosis. Strong membranous immunoreactivity of FGFR1 in glandular epithelial cells of the ectopic endometrium of endometriosis was observed compared with their eutopic counterparts (Fig. 2C and D). These findings are regardless of either proliferative
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
121
Fig. 2. FGFR1 overexpression in endometriosis. (A) Representative images of Western blotting analysis of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium obtained by LCM. b-Actin was used as a loading control. (B) Box plot showing the differential expression of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. (C) Representative images of immunohistochemical analysis of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. Dark yellow to brown color represents positive signal of FGFR1. (D) Box plot showing the differential expression of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. NC, negative control. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
phase or secretory phase (P > 0.05). These observations suggested that FGFR1 might be correlated with endometriosis formation. Abrogation of endometriotic lesion in nude mice by FGFR1 inhibition We further examined whether FGFR1 is critical for endometriosis formation in vivo. The incidence of endometriotic nodules in PBS-, Lipo- or FGFR1 shRNA-treated endometriosis nude mice model at the time of sacrifice was listed in Table 2. In mice treated with PBS and Lipo, measurable endometriotic tissues after a 25-day incubation were observed (Fig. 3A), with an average volume of 32.2 and 29.8 mm3, respectively (Fig. 3D). By contrast, endometriotic tissues in mice treated with FGFR1 shRNA could hardly be detectable after 25 days of growth (Fig. 3A and D). The average volume of endometriotic nodules in this group was only 12.0 mm3 at sacrifice (Fig. 3D). Moreover, the average weight of endometriotic tissues in the FGFR1 shRNA-treated group was significantly lower than those in either PBS- or FGFR1 shRNA-treated group (P = 0.0017; Fig. 3E). At Table 2 Summary of the incidence of endometriotic nodules in PBS-, Lipo- or FGFR1 shRNAtreated endometriosis nude mice model at the time of sacrifice. Mouse Tx
PBS Lipo FGFR1 shRNA
No of mice with disease per total number (%)
No of foci (mean SD)
5/5 (100%) 5/5 (100%) 3/5 (60%)
3.4 1.14 3.6 1.34 0.6 0.55
the time of sacrifice, mice treated with both PBS and Lipo showed persistence of active lesions with angiogenic and glandular organization (the scores were 2.2 0.76 and 2.3 0.57, respectively; Fig. 3B). By contrast, among the five nude mice treated with FGFR1 shRNA, two mice (40%) showed complete regression of endometriotic lesions, with the remaining three mice displaying fibrotic and avascular lesions (score 0.5 0.5; Fig. 3B). The pathology scores of the mice treated with FGFR1 shRNA and the mice in the two control groups demonstrated significant difference (P = 0.0009; Fig. 3B). In addition, as He et al. previously demonstrated that dysmenorrhea in endometriosis patients derives from generalized hyperalgesia [14], we next examined whether FGFR1 knockdown could reduce generalized hyperalgesia in endometriosis model in vivo. Nude mice that received FGFR1 shRNA treatment demonstrated a significant improvement in hot plate latency (P < 0.0001; Fig. 3F), indicating that FGFR1 knockdown could alleviate endometriosis-induced hyperalgesia in nude mice. Further immunohistochemistry analysis for markers of proliferation (Ki67) in the three groups demonstrated decreased proliferation of the endometriotic lesions treated with FGFR1 shRNA, compared with those in the PBS and Lipo groups (Fig. 3C). These observations proved that FGFR1 inhibition could effectively abrogate xenotransplanted endometriotic lesion growth and proliferation in nude mice. Ectopic FGFR1 expression was correlated with severity of dysmenorrhea and recurrence in endometriosis We further examined whether FGFR1 expression in endometrium is correlated with patient rAFS score and stage and found
122
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
Fig. 3. Effects of FGFR1 inhibition on experimental endometriosis model in nude mice. (A) Representative images of PBS-, Lipo-, or FGFR1 shRNA-treated endometriosis nude mice model. The black arrows indicate endometriotic lesion sites. On the right, the number of endometriotic nodules was quantified in endometriosis nude mice model in each group (n = 5 per group). (B) Representative images of serial hematoxylin and eosin (H&E) staining of endometriotic lesion sections. On the right is the pathology scores of experimental endometriosis in nude mice of each group. Specimens of experimental endometriosis model of nude mice in each group were immunostained for Ki-67. On the right is the scatter plot of expression level of Ki-67 in the experimental endometriosis nude mice model in each group (n = 5 per group). (D) The lesion volume of endometriotic nodules was measured in endometriosis nude mice model in each group 25 days postinoculation (n = 5 per group). (E) The weight of endometriotic nodules was quantified in endometriosis nude mice model in each group 25 days postinoculation (n = 5 per group). (F) Time course of changes in average hot plate latency in respective groups. The abbreviated words in the figure represent different time points: Tx stands for treatment, and Exp’t stands for experiment. All of the data are from at least three independent experiments and are shown as the means S.D. *P < 0.05; **P < 0.01; ***P < 0.001.
that no relationship between both ectopic and eutopic endometrial FGFR1 immunoreactivity levels and rAFS stage and score (data not shown). We then evaluated the relationship between FGFR1 expression in endometrium and dysmenorrhea severity in these patients. The baseline characteristics of the recurrence and nonrecurrence groups between the two groups were demonstrated in Table 3. Results showed that while eutopic FGFR1 expression demonstrated no significant significance among endometriosis patients without or with mild, moderate or severe dysmenorrhea (Fig. 4A), FGFR1 expression in ectopic endometrium was significantly higher in endometriosis patients who reported severe dysmenorrhea than those reporting no, mild or moderate dysmenorrhea (P < 0.05, Fig. 4B). Pearson’s correlation analysis identified a positive correlation between FGFR1 expression
(adjusted by menstrual phase) in ectopic endometrium and dysmenorrhea scores in 48 endometriosis patients with statistical significance (r2 = 0.5223; P < 0.0001, Pearson’s x2 test; Fig. 4C). Moreover, the immunoreactivity level of ectopic FGFR1 expression in the recurrence group was significantly higher than that in the control and non-recurrent group (P < 0.05, Fig. 4D). Collectively, our results implied that ectopic FGFR1 expression is correlated with dysmenorrhea severity and recurrence in endometriosis patients. Comments In this study, we used a novel bioinformatics-based strategy to screen for molecules responsible for endometriosis pathogenesis
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
123
Table 3 Baseline characteristics of the recurrence and non-recurrence groups between the two groups. Variable
Recurrence group (n = 20)
Non-recurrence group (n = 28)
P value
Age Sampling time during menstrual cycle Proliferative Secretory Body mass index Previous endometriosis-related surgery No Yes Complaint of infertility No Yes Size
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Increased expression of fibroblast growth factor receptor 1 in endometriosis and its correlation with endometriosis-related dysmenorrhea and recurrence Linjie Zhao a,b, Huiliang Yang a,b, Yu Xuan a,b, Zhongyue Luo c, Qiao Lin c, Jitong Zhao a,b, Ning Ren a,b, Shengtao Zhou a,b,*, Xia Zhao a,b a
Department of Gynecology and Obstetrics, Key Laboratory of Obstetrics & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second Hospital, Chengdu 610041, PR China b The State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, Chengdu 610041, PR China c College of Life Science, Sichuan University, Chengdu 610041, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 June 2014 Received in revised form 23 October 2014 Accepted 11 November 2014
Objective(s): This study aims to identify a critical molecule that potentially participates in endometriosis pathogenesis and characterize its correlation with dysmenorrhea and recurrence. Study design: We utilized a bioinformatics-based strategy to screen for candidate genes and fibroblast growth factor receptor 1(FGFR1) was chosen for further validation. FGFR1 expression was examined in specimens of ectopic and eutopic endometrium obtained from 48 patients with endometriosis and specimens of eutopic endometrium from 26 healthy control subjects using immunohistochemistry and Western blotting. In addition, FGFR shRNA treatment was applied in a nude mice endometriosis model to examine the functional role of FGFR1 in endometriosis formation in vivo. Results: FGFR1 was found commonly overexpressed in ectopic endometrium of endometriosis compared with either its eutopic counterpart or endometrium from normal patients (P < 0.05). FGFR shRNA treatment impaired endometriosis formation and alleviated endometriosis-related symptoms in vivo. FGFR1 expression in ectopic endometrium was correlated with dysmenorrhea severity (P < 0.05) and recurrence in endometriosis patients (P < 0.05). Conclusion(s): FGFR1 might be involved in endometriosis development, which could possibly serve as a novel therapeutic target and prognostic marker for this disease. ß 2014 Published by Elsevier Ireland Ltd.
Keywords: Endometriosis FGFR1 Dysmenorrhea Recurrence
Introduction Endometriosis is a common estrogen-dependent gynecological disorder in females of reproductive age and is characterized by the presence and growth of endometrial tissue outside the uterus [1]. The development of ectopic endometrial tissues suggests intrinsic cellular mechanisms leading to invasion, unrestrained growth, neoangiogenesis, and distant spreading of endometriotic cells [2]. Despite its high prevalence, the etiology and pathogenesis of endometriosis still remain largely unelucidated. Moreover, although hormonal therapy offers an alternative for the treatment of endometriosis, its curative efficacy for endometriosis is often
* Corresponding author at: Department of Gynecology and Obstetrics, West China Second Hospital, Sichuan University, Chengdu, 610041, PR China. Tel.: +86 28 13551070137.. E-mail address: [email protected] (S. Zhou). http://dx.doi.org/10.1016/j.ejogrb.2014.11.013 0301-2115/ß 2014 Published by Elsevier Ireland Ltd.
insufficient. Therefore, the development of novel treatment strategies for endometriosis based upon new potential therapeutic targets is imperative. Previous observations supported the notion that endometriosis might be a ‘‘metastasis’’-related disease, which offers endometriosis some similarities to cancer diseases although metastasis does not really exist in endometriosis. However, the underlying mechanisms are still unclear [3,4]. Fibroblast growth factor receptor (FGFR) family is part of a group of growth factor receptor tyrosine kinases (RTK) able to induce several cellular processes including cell proliferation, angiogenesis, inhibition of apoptosis, and cell migration [5]. FGFR1, as a member of fibroblast growth factor receptors, is known to drive an epithelial-to-mesenchymal transition (EMT) of primitive streak-localized epiblast cells into mesoderm cells. During EMT, epithelial cells lose their polarity, augment their motility, and begin to express mesenchymal markers, such as vimentin, becoming ‘‘mesenchymal-like’’. This process has also been closely linked to cancer progression and
118
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
metastasis. Previously, aberrant FGFR1 expression has been implicated in the progression of a variety of tumor cell types including prostate cancer, colorectal cancer and non-small lung cancer [5]. However, the contribution of FGFR1 to the pathogenesis of benign disorders like endometriosis is still unclear. Thus, our study attempted to identify a critical molecule that might participate in endometriosis development. Materials and methods Bioinformatics analysis Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI) was searched to identify candidate genes responsible for endometriosis development. Three GEO datasets analyzing endometriosis were identified, including GSE5108, GSE6364, and GSE7305 [6–8]. These three datasets were compared for overlapped genes differentiallyexpressed between endometriosis and control. Two criteria were applied: (a) significantly differentially-regulated (P < 0.05) between endometriosis and control identified by GEO, (b) significantly differentially-expressed by at least 2-fold (P < 0.05). Since endometriosis might be a ‘‘metastasis’’-related disease similar to cancer although metastasis does not really exist, we further searched eligible genes against KEGG pathway database, Gene Ontology and metastasis-related published studies to identify genes most closely associated with metastasis. These ‘‘metastasis’’ genes were subsequently analyzed in STRING software, an open source web-based tool with established and predicted protein interactions, and visualized protein-protein interaction networks [9]. Those key nodes with over two interactive nodes demonstrated by STRING pathway analysis were selected as potential candidate genes responsible for the pathogenesis of endometriosis. Patients and tissue samples Eutopic endometrium used for Western blot analysis was obtained from 26 patients undergoing hysterectomy for benign indications other than endometriosis (like leiomyoma) as control, and ectopic endometrium was obtained from endometriosis patients at the Gynecological Department of West China Second Hospital of Sichuan University (Chengdu, China) from 2009 to 2014. None of the patients had received any preoperative hormonal therapy prior to surgery. All these samples were obtained by experienced gynecologists and gynecological surgeons and examined by experienced pathologists who confirmed the diagnosis of disease samples. Paraffin-embedded eutopic endometrial and paired ectopic endometrial specimens were obtained from 48 patients who underwent surgical resections in the same hospital from 2009 to 2014. They were further subjected to immunohistochemical analysis of identified proteins for validation and subsequent follow-up were carried out for these 48 patients. This study was approved by the Institutional Ethics Committee of Sichuan University. Informed consents were obtained from all patients prior to analysis. Laser capture microdissection (LCM) and Western blotting All lesions were microdissected from 12-mm-thick frozen sections using LCM system (Leica, Germany). Normal endometrial tissues were dissected to obtain negative control proteins. Western blotting analysis was conducted as previously described [10]. The primary antibody was rabbit anti-FGFR1 (Santa Cruz Biotechnology). b-Actin was used as an internal control.
(Santa Cruz Biotechnology) and rabbit anti-Ki-67 (Santa Cruz Biotechnology). Slides were evaluated by two independent pathologists in a double-blinded manner. Any discrepancy between the two evaluators was resolved by reevaluation and careful discussion until agreement was reached. Animal model of endometriosis The guidelines for animal care were approved by the Institutional Animal Care and Use Committee of Sichuan University (Chengdu, Sichuan, People’s Republic of China). Endometriosis lesions were acquired from premenopausal women with endometriosis at West China Second Hospital, Sichuan University. In vivo endometriosis nude mice model was surgically induced as described previously [11]. In particular, each mouse received an intraperitoneal injection of PBS containing a suspension of five human endometriotic tissue fragments with a size of about 10 mm3 per mouse into the ventral midline and subcutaneous injection of 0.5 mg of 17b-estradiol was performed on days 1 and 2 to facilitate the implantation of endometriotic lesions. Mice were next assigned randomly to one of the following groups (5 per group): (a) PBS, 100 ml of PBS; (b) Lipo, Lipofectamine 2000 (Invitrogen) at 62.5 mg/100 ml of PBS; (c) shFGFR1, FGFR1 shRNA at 25 mg/100 ml of PBS. Intraperitoneal treatment was initiated 5 days after inoculation. Mice received therapy, three times a week and were sacrificed at 21 days postinoculation. Intraperitoneal endometrial nodules were resected and measured immediately to assess the treatment efficacy as previously described [10]. Pain assessment and follow-up of endometriosis patients Pain assessment of endometriosis patients was conducted as described elsewhere [12]. The 48 patients with endometriosis were further followed up for recurrence for 30 months after surgery. The recurrence of endometriosis was defined as: (a) the presence of ovarian cysts of 3 cm in diameter, along with characteristic echoes as detected by transvaginal ultrasonography for two consecutive menstrual cycles, coupled with or without the recurrence of dysmenorrhea or pelvic pain requiring medical intervention, or as (b) the presence of de novo endometriosis as confirmed by histology following a second surgery. We also included endometrial tissue samples from 26 women with other benign gynecological diseases as controls and none of them had endometriosis or adenomyosis. Statistics Data are presented as mean SD of three independent experiments unless otherwise indicated. GraphPad Prism (GraphPad Software Inc., CA, USA) was used for data analysis. The comparison of distributions of continuous variables was made using the Wilcoxon test or Kruskal–Wallis test. Jonckheere–Terpstra trend test was utilized to test for trend of FGFR1 immunoreactivity in women reporting more severe dysmenorrhea as described previously [13]. The correlation between FGFR1 staining scores and dysmenorrhea scores was analyzed using Pearson’s x2 test. Comparisons between two groups were performed with the Student’s t test, and differences among multiple groups were evaluated by one-way ANOVA analysis. Differences were considered statistically significant at P < 0.05. Results
Immunohistochemistry
Bioinformatics identification of FGFR1 as a candidate gene for endometriosis development
Immunohistochemistry (IHC) was performed as described previously [10]. The primary antibodies included rabbit anti-FGFR1
A total of 493 genes significantly differentially-regulated by at least 2-fold (P < 0.05) and overlapped in three datasets were
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
identified (Fig. 1A and B). Among the 493 overlapped genes, 43 genes were further identified to be ‘‘metastasis-related’’. The detailed information of each identified gene was listed in Table 1 and this gene repertoire fell into distinct categories based upon their sub-cellular localizations (Fig. 1C) and biological functions (Fig. 1D). Subsequently, these genes were further subjected to
119
STRING pathway analysis and an interaction network was visualized (Fig. 1E). Based upon the above bioinformatics analysis, FGFR1 was selected for further functional study due to (a) the unclear role of FGFR1 in the pathogenesis of endometriosis; (b) its ‘‘plasma’’ membrane location, which facilitates it to sense extracellular
Fig. 1. Bioinformatics analysis of differentially expressed genes in ectopic and eutopic endometrium of human endometriosis. (A) Screening strategy for key molecules for endometriosis pathogenesis. (B) Venn diagram for overlapped genes of the three publicly available endometriosis datasets. (C) Gene expression cluster map generated by Cluster software. Expression of genes in the eutopic endometrium is constant at 0, whereas genes up-regulated in the ectopic endometrium are in red, and the downregulated proteins are in green. The intensity of the color green or red corresponds to the degree of alteration, respectively, according to the color strip at the bottom of the figure. (D) The identified genes were categorized into several protein groups according to their subcellular locations. (E) The identified genes were categorized into several protein groups according to their biological functions. (F) The protein-protein interaction network of identified genes analyzed by STRING software. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
120
Table 1 Overlapped metastasis-related genes in the three publicly available datasets. Gene no.
Accession no.
Protein name
Gene name
Average alteration in three datasets (fold)
Subcellular location
Function
1 2 3
P14651 Q12802 P63241
HOXB3 AKAP13 EIF5A
6.255 1.4 1.26
Nucleus Cytoplasm Cytoplasm
Angiogenesis Apoptosis Apoptosis
4 5 6 7
P40240 Q75L79 P43121 Q8IZT6
CD9 CLDN3 MCAM ASPM
1.41 6.87 1.68 4.655
Membrane Membrane Membrane Cytoplasm
Cell Cell Cell Cell
8 9
P48729 P23921
1.505 2.245
Cytoplasm Cytoplasm
Cell cycle Cell cycle
10 11 12 13 14 15 16 17 18
O14818 Q96GD4 Q9UNL4 P15923 P62826 P49815 P16949 Q8WVJ9 P25490
Cytoplasm Nucleus Nucleus Nucleus Nucleus Cytoplasm Cytoplasm Nucleus Nucleus
Cell Cell Cell Cell Cell Cell Cell Cell Cell
19
O75030
20 21
Q8NI77 O60271
22 23 24 25 26 27 28 29 30
P14923 P42345 P04626 P21246 P11362 P57735 P03372 Q9BXP5 P67809
31
Q9BY41
Homeobox protein Hox-B3 A-kinase anchor protein 13 Eukaryotic translation initiation factor 5A-1 CD9 antigen Claudin 3 Cell surface glycoprotein MUC18 Abnormal spindle-like microcephaly-associated protein Casein kinase I isoform alpha Ribonucleoside-diphosphate reductase large subunit Proteasome subunit alpha type-7 Aurora kinase B Inhibitor of growth protein 4 Transcription factor E2-alpha GTP-binding nuclear protein Ran Tuberin Stathmin Twist-related protein 2 Transcriptional repressor protein YY1 Microphthalmia-associated transcription factor Kinesin-like protein KIF18A C-Jun-amino-terminal kinase-interacting protein 4 Junction plakoglobin Serine/threonine-protein kinase mTOR Receptor tyrosine-protein kinase erbB-2 Pleiotrophin Fibroblast growth factor receptor 1 Ras-related protein Rab-25 Estrogen receptor Serrate RNA effector molecule homolog Nuclease-sensitive element-binding protein 1 Histone deacetylase 8
32
P24347
Stromelysin-3
MMP11
4.53
Secreted
33
P19235
Erythropoietin receptor
EPOR
1.54
Membrane
34 35 36 37
Q06455 B0YJ81 P00374 Q9HCC0
RUNX1T1 PTPLA DHFR MCCC2
38
Q16630
39
P62310
40
Q15599
41
Q92766
42
O76094
43
P05162
Protein CBFA2T1 3-hydroxyacyl-CoA dehydratase 1 Dihydrofolate reductase Methylcrotonoyl-CoA carboxylase beta chain, mitochondrial Cleavage and polyadenylation specificity factor subunit 6 U6 snRNA-associated Sm-like protein LSm3 Na(+)/H(+) exchange regulatory cofactor NHE-RF2 Ras-responsive element-binding protein 1 Signal recognition particle 72 kDa protein Galectin-2
signals and conduct information into individual cells; (c) its function in mediating cell migration under physiological conditions and tumor cell metastasis. Overexpression of FGFR1 in the ectopic endometrium of endometriosis To determine FGFR1 expression in human endometriotic tissues, we performed LCM to obtain endometrial tissues of paired eutopic and ectopic endometrium of endometriosis patients and
CSNK1A1 RRM1 PSMA7 AURKB ING4 TCF3 RAN TSC2 STMN1 TWIST2 YY1
1.24 3.24 1.13 1.94 1.64 1.26 2.03 3.05 1.16
adhesion adhesion adhesion cycle
cycle cycle cycle cycle cycle cycle differentiation differentiation differentiation
MITF
1.885
Nucleus
Cell differentiation
KIF18A SPAG9
3.205 1.395
Cytoplasm Cytoplasm
Cell migration Cell migration
JUP MTOR ERBB2 PTN FGFR1 RAB25 ESR1 SRRT YBX1
2.725 1.29 1.255 2.51 1.616667 4.37333 5.26 1.3 1.13
Membrane Cytoplasm Cytoplasm Secreted Membrane Membrane Nucleus Nucleus Cytoplasm
Cell Cell Cell Cell Cell Cell Cell Cell Cell
HDAC8
1.71
Nucleus
Nucleus Cytoplasm Cytoplasm Cytoplasm
Chromatin assembly or disassembly Collagen catabolic process Erythropoietin receptor activity Metabolism Metabolism Metabolism Metabolism
1.53 1.84 1.86 1.95333
migration proliferation proliferation proliferation proliferation proliferation proliferation proliferation proliferation
CPSF6
1.36333
Nucleus
mRNA processing
LSM3
1.32
Nucleus
mRNA processing
SIP1
1.38
Membrane
RREB1
1.295
Nucleus
Protein complex assembly Signal transduction
SRP72
1.32
Cytoplasm
Signal transduction
2.87667
Cytoplasm
Un known
LGALS2
control for further Western blotting analysis. Elevated FGFR1 expression was detected in ectopic endometrium of endometriosis patients compared with normal endometrium and eutopic endometrium of endometriosis (Fig. 2A and B). In addition, FGFR1 expression was validated in paraffin-embedded specimens of endometriosis. Strong membranous immunoreactivity of FGFR1 in glandular epithelial cells of the ectopic endometrium of endometriosis was observed compared with their eutopic counterparts (Fig. 2C and D). These findings are regardless of either proliferative
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
121
Fig. 2. FGFR1 overexpression in endometriosis. (A) Representative images of Western blotting analysis of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium obtained by LCM. b-Actin was used as a loading control. (B) Box plot showing the differential expression of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. (C) Representative images of immunohistochemical analysis of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. Dark yellow to brown color represents positive signal of FGFR1. (D) Box plot showing the differential expression of FGFR1 in ectopic and paired eutopic endometrium of endometriosis patients and normal endometrium. NC, negative control. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
phase or secretory phase (P > 0.05). These observations suggested that FGFR1 might be correlated with endometriosis formation. Abrogation of endometriotic lesion in nude mice by FGFR1 inhibition We further examined whether FGFR1 is critical for endometriosis formation in vivo. The incidence of endometriotic nodules in PBS-, Lipo- or FGFR1 shRNA-treated endometriosis nude mice model at the time of sacrifice was listed in Table 2. In mice treated with PBS and Lipo, measurable endometriotic tissues after a 25-day incubation were observed (Fig. 3A), with an average volume of 32.2 and 29.8 mm3, respectively (Fig. 3D). By contrast, endometriotic tissues in mice treated with FGFR1 shRNA could hardly be detectable after 25 days of growth (Fig. 3A and D). The average volume of endometriotic nodules in this group was only 12.0 mm3 at sacrifice (Fig. 3D). Moreover, the average weight of endometriotic tissues in the FGFR1 shRNA-treated group was significantly lower than those in either PBS- or FGFR1 shRNA-treated group (P = 0.0017; Fig. 3E). At Table 2 Summary of the incidence of endometriotic nodules in PBS-, Lipo- or FGFR1 shRNAtreated endometriosis nude mice model at the time of sacrifice. Mouse Tx
PBS Lipo FGFR1 shRNA
No of mice with disease per total number (%)
No of foci (mean SD)
5/5 (100%) 5/5 (100%) 3/5 (60%)
3.4 1.14 3.6 1.34 0.6 0.55
the time of sacrifice, mice treated with both PBS and Lipo showed persistence of active lesions with angiogenic and glandular organization (the scores were 2.2 0.76 and 2.3 0.57, respectively; Fig. 3B). By contrast, among the five nude mice treated with FGFR1 shRNA, two mice (40%) showed complete regression of endometriotic lesions, with the remaining three mice displaying fibrotic and avascular lesions (score 0.5 0.5; Fig. 3B). The pathology scores of the mice treated with FGFR1 shRNA and the mice in the two control groups demonstrated significant difference (P = 0.0009; Fig. 3B). In addition, as He et al. previously demonstrated that dysmenorrhea in endometriosis patients derives from generalized hyperalgesia [14], we next examined whether FGFR1 knockdown could reduce generalized hyperalgesia in endometriosis model in vivo. Nude mice that received FGFR1 shRNA treatment demonstrated a significant improvement in hot plate latency (P < 0.0001; Fig. 3F), indicating that FGFR1 knockdown could alleviate endometriosis-induced hyperalgesia in nude mice. Further immunohistochemistry analysis for markers of proliferation (Ki67) in the three groups demonstrated decreased proliferation of the endometriotic lesions treated with FGFR1 shRNA, compared with those in the PBS and Lipo groups (Fig. 3C). These observations proved that FGFR1 inhibition could effectively abrogate xenotransplanted endometriotic lesion growth and proliferation in nude mice. Ectopic FGFR1 expression was correlated with severity of dysmenorrhea and recurrence in endometriosis We further examined whether FGFR1 expression in endometrium is correlated with patient rAFS score and stage and found
122
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
Fig. 3. Effects of FGFR1 inhibition on experimental endometriosis model in nude mice. (A) Representative images of PBS-, Lipo-, or FGFR1 shRNA-treated endometriosis nude mice model. The black arrows indicate endometriotic lesion sites. On the right, the number of endometriotic nodules was quantified in endometriosis nude mice model in each group (n = 5 per group). (B) Representative images of serial hematoxylin and eosin (H&E) staining of endometriotic lesion sections. On the right is the pathology scores of experimental endometriosis in nude mice of each group. Specimens of experimental endometriosis model of nude mice in each group were immunostained for Ki-67. On the right is the scatter plot of expression level of Ki-67 in the experimental endometriosis nude mice model in each group (n = 5 per group). (D) The lesion volume of endometriotic nodules was measured in endometriosis nude mice model in each group 25 days postinoculation (n = 5 per group). (E) The weight of endometriotic nodules was quantified in endometriosis nude mice model in each group 25 days postinoculation (n = 5 per group). (F) Time course of changes in average hot plate latency in respective groups. The abbreviated words in the figure represent different time points: Tx stands for treatment, and Exp’t stands for experiment. All of the data are from at least three independent experiments and are shown as the means S.D. *P < 0.05; **P < 0.01; ***P < 0.001.
that no relationship between both ectopic and eutopic endometrial FGFR1 immunoreactivity levels and rAFS stage and score (data not shown). We then evaluated the relationship between FGFR1 expression in endometrium and dysmenorrhea severity in these patients. The baseline characteristics of the recurrence and nonrecurrence groups between the two groups were demonstrated in Table 3. Results showed that while eutopic FGFR1 expression demonstrated no significant significance among endometriosis patients without or with mild, moderate or severe dysmenorrhea (Fig. 4A), FGFR1 expression in ectopic endometrium was significantly higher in endometriosis patients who reported severe dysmenorrhea than those reporting no, mild or moderate dysmenorrhea (P < 0.05, Fig. 4B). Pearson’s correlation analysis identified a positive correlation between FGFR1 expression
(adjusted by menstrual phase) in ectopic endometrium and dysmenorrhea scores in 48 endometriosis patients with statistical significance (r2 = 0.5223; P < 0.0001, Pearson’s x2 test; Fig. 4C). Moreover, the immunoreactivity level of ectopic FGFR1 expression in the recurrence group was significantly higher than that in the control and non-recurrent group (P < 0.05, Fig. 4D). Collectively, our results implied that ectopic FGFR1 expression is correlated with dysmenorrhea severity and recurrence in endometriosis patients. Comments In this study, we used a novel bioinformatics-based strategy to screen for molecules responsible for endometriosis pathogenesis
L. Zhao et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 184 (2015) 117–124
123
Table 3 Baseline characteristics of the recurrence and non-recurrence groups between the two groups. Variable
Recurrence group (n = 20)
Non-recurrence group (n = 28)
P value
Age Sampling time during menstrual cycle Proliferative Secretory Body mass index Previous endometriosis-related surgery No Yes Complaint of infertility No Yes Size
