Protease-activated receptor 2 plays an important role in the pathogenesis of endometriosis. We studied the effect of ENMD-1068, a protease-activated receptor 2 antagonist, on the development of endometriosis in a noninvasive fluorescent mouse model.
A red fluorescent protein–expressing xenograft model of human endometriosis was created in nude mice. After endometriosis induction, the mice were injected intraperitoneally with either 25 mg/kg or 50 mg/kg ENMD-1068 or with 200 μL of the vehicle control daily for 5 days. The endometriotic lesions that developed in the mice were then counted, measured, and collected. The lesions were assessed for the production of interleukin 6 and monocyte chemotactic protein-1 by enzyme-linked immunosorbent assays and evaluated for the activation of nuclear factor-κB and the expression of vascular endothelial growth factor by immunohistochemical analyses. Cell proliferation and apoptosis were assessed by immunohistochemistry for Ki-67 and terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling, respectively.
ENMD-1068 dose-dependently inhibited the development of endometriotic lesions ( P < .05) without apparent toxicity to various organs of the treated mice. Consistently, ENMD-1068 dose-dependently inhibited the expression of interleukin 6 and nuclear factor-κB ( P < .05) and cell proliferation ( P < .05) in the lesions, as well as increased the percentage of apoptotic cells ( P < .05). ENMD-1068 reduced the levels of monocyte chemotactic protein-1 and vascular endothelial growth factor in the lesions ( P < .05), but not in a dose-dependent manner.
Our study suggests that ENMD-1068 is effective in suppressing the growth of endometriosis, which might be attributed to the drug’s antiangiogenic and antiinflammatory activities.
Endometriosis is a benign gynecological disease characterized by the presence and growth of endometrial tissue outside the uterus. It affects approximately 6-10% of reproductive women and mainly results in chronic pelvic pain, dysmenorrhea, and infertility. Moreover, this disorder could increase the risk of ovarian cancer. Current treatment of endometriosis involves surgery or medication. However, symptoms recur in up to 75% of women within 2 years of surgical treatment. Drug therapies also have side effects that limit their long-term use. The identification of new therapeutic targets, possibly from genomic and proteomic analyses, is needed to develop more useful treatments.
Protease-activated receptor 2 (PAR2), a G-protein-coupled receptor activated by serine proteinases, has been implicated in the development of endometriosis. Hirota et al and Osuga et al demonstrated that PAR2 activation could stimulate the proliferation of endometriotic stromal cells and induce an increase in interleukin (IL)-6 and IL-8 secretion by these cells. Moreover, the number and total weight of endometriotic lesions were significantly decreased and inflammatory substance release from endometriotic lesions was attenuated in PAR2 knockout mice, arguing for a critical proinflammatory role for this receptor in endometriosis.
Recently, a study from our group showed a significant increase in gene and protein expression of tissue factor (TF) and PAR2 in eutopic and ectopic endometrium with endometriosis. We speculated that TF and PAR2 may be involved in the pathogenesis of the disease. TF or PAR2 would be an ideal target for the treatment of endometriosis.
However, TF is the principal initiator of blood clotting, and thus targeting TF runs the risk of coagulation disorder or dysfunction. A clinically relevant question that can be addressed is whether PAR2 presents a new therapeutic target for the treatment of endometriosis. ENMD-1068 is a novel selective antagonist of PAR2 and is able to prevent PAR2 activation from genomic intervention. In the present study, a mouse model of endometriosis with red fluorescent protein (RFP) expression for noninvasive visualization was used to establish a proof of concept, assessing the effect of ENMD-1068 on the development of endometriosis. The effect of ENMD-1068 treatment in the mouse model was evaluated in terms of endometriotic lesion establishment, cell proliferation, and apoptosis of ectopic endometrial tissue. The release of proinflammatory factors (IL-6 and monocyte chemotactic protein [MCP]-1), activation of nuclear factor (NF)-κB, and expression of vascular endothelial growth factor (VEGF) in the lesions were also analyzed.
Materials and Methods
Acquisition of endometrial tissues
The diagnosis of endometriosis was confirmed by laparotomy or laparoscopy and pathology (when biopsy was performed). Eutopic endometrial tissues were acquired through curettage during the secretory phase (days 15-28) of the normal menstrual cycle from 18 ovarian endometrioma volunteer donors (aged 31.2 ± 5.1 years). Individuals with a recent (within 6 months) history of hormone therapy (ie, oral contraceptives) were excluded. Endometrial biopsies were immediately transferred into RPMI 1640 medium supplement with 2% fetal calf serum (FCS), 100 μg/mL penicillin, and 100 IU streptomycin (Wisent Inc, Saint Bruno, Quebec, Canada). All patients gave their written informed consent before tissue collection, and the research protocol was approved by the institutional ethical and review board of Zhujiang Hospital of Southern Medical University, Guangzhou, China.
Preparation of endometrial fragments and transduction with RFP
Human endometrial tissue was washed with phosphate-buffered saline (PBS), transferred to DMEM/F12 (Sigma, St. Louis, MO) supplemented with 10% FCS and 10 nmol/L 17β-estradiol (E2) (Sigma), and cut into 1–2 mm 3 pieces. Six tissue fragments were distributed into each well of a 48-well plate. Transduction of the endometrial fragments was performed by incubation at 37°C, 5% carbon dioxide, with lentiviral vector particles encoding RFP (GeneChem, Shanghai, China) at a multiplicity of infection of 500. After a 24-hour incubation with the lentiviral vector particles, the endometrial fragments were harvested, washed 3 times with PBS, and resuspended in PBS with 2% FCS at a concentration of 10-12 tissue fragments per 500 μL.
Surgical induction of endometriosis
The guidelines for animal welfare were approved by the ethics committee on animal experimental of the Zhongshan Ophthalmic Center, Sun Yat-sen University (Guangzhou, China). Twenty-five 6- to 8-week-old ovariectomized female nude mice (nude/nu) were purchased from Animal Experimental Center, Guangdong Academy of Medical Science (Guangzhou, China). After a 1-week acclimation period, mice were anesthetized with isoflurane. Human endometrial tissue fragments were placed by laparotomy into the peritoneal cavity of mice in the region of gut, lateral abdominal wall, liver, and mesenterium via a single 5-mm ventral incision. Each mouse received an intraperitoneal implantation of 10-12 RFP-expressing endometrial tissue fragments. At the time of surgery, all mice were implanted subcutaneously with a silastic capsule containing 8 mg of 17β-estradiol (E2) in cholesterol (Sigma-Aldrich, St. Louis, MO). Eight days after the tissue implantation, the noninvasive imaging of RFP expression in the mice was conducted with a NightOWL LB 983 in vivo optical imaging system (Berthold Technologies, Bad Wildbad, Germany). For the imaging, mice were anesthetized with isoflurane, placed in the light-tight chamber, and illuminated with the filters for excitation and emission set at 570 nm and 600 nm, respectively. A gray-scale reference image was collected under white light, and it was merged with pseudocolor using WinLight software (Berthold Technologies) to localize the photon emission of endometriotic lesions in the mice. From the fluorescence images, the peak and integrated bioluminescence fluxes (ph/s) were calculated with the software. If RFP could not be seen inside the endometriotic lesions, the mice were subjected to a laparotomy to allow the identification of the lesions. The RFP fluorescence or macroscopic lesions that were seen in the mice showed the growth of human endometrial tissue in nude mice.
Treatment was initiated 10 days after the implantation of endometrial tissue and was continued for 5 days. Mice with surgically induced endometriosis were randomly assigned to 3 experimental groups (n = 8 per group): 25 mg/kg ENMD-1068 (Enzo Life Sciences Inc, Farmingdale, NY), 50 mg/kg ENMD-1068, and saline solution control (200 μL). Each type of treatment was administered once per day by intraperitoneal injection. All the mice were monitored daily, and any evidence of toxicity at each treatment dose administered was noted based on body weight, food consumption, grooming behavior, or activity levels.
Evaluation of endometriotic lesions
After 5 days of treatment, animals were sacrificed by cervical dislocation. Lesions were identified, counted, and measured using a sterile caliper. The average volume of the lesions was calculated by the formula V = a × 2b × 0.5, where “a” and “b” are the largest and the smallest superficial diameters of the lesion, respectively. After the wet weight of the lesions per mouse was measured, the lesions were prepared for enzyme-linked immunosorbent assays (ELISA).
ELISA for IL-6 and MCP-1
To investigate cytokine production by the endometrial allografts, ectopic lesions from each mouse were immediately incubated in 0.5 mL of sterile M199 culture medium (pH 7.4) for 2 hours at 37°C. After centrifugation at 3000 g and 4°C for 20 minutes, ELISA kits (Cusabio Biotech Co Ltd, Wuhan, Hubei Province, China) were used to measure the concentrations of IL-6 and MCP-1 in the supernatant according to the manufacturer’s instructions. The results were expressed as the amount produced per unit wet weight of tissue per 2 hours (pg/mg/2 h). Then the tissues were fixed in 10% formaldehyde for histological evaluation, immunohistochemistry, and cell death determination.
Immunohistochemistry for NF-κB and Ki-67
Sections (4-μm thick) of endometrial tissue were deparaffinized in xylol and rehydrated through graded alcohols. For antigen retrieval, sections were microwaved for 5 minutes at 600 W in 0.01 mmol/L sodium citrate buffer (pH 6.0). After rinsing in PBS (pH 7.4), the slides were immersed in 3% H 2 O 2 for 15 minutes to inhibit endogenous peroxidase activity. The sections were then rinsed in PBS and incubated at room temperature with primary antibodies against the NF-κB p65 (RelA) subunit (1:50; Santa Cruz Biotechnology, Santa Cruz, CA), VEGF (1:100; Santa Cruz Biotechnology), and Ki-67 (1:100; BD, Franklin Lakes, NJ) for 60 minutes, and rinsed in PBS. Normal serum was used instead of primary antibodies as negative controls. The sections were thereafter incubated with biotinylated antimouse or antirabbit secondary antibody for 30 minutes. Antibody-bound sites were visualized by avidin-biotin-peroxidase complex solution and 3, 3′-diaminobenzidine as a chromogen.
The immunoreactivity scores of NF-κB p65 (RelA) and VEGF were evaluated by digital image analysis using Image-Pro Plus 6.0 (Media Cybernetics, Bethesda, MD). Basically, 10 nonoverlapping, randomly chosen images of the sections were taken at ×400 magnification with a digital camera (Olympus DP70; Olympus, Tokyo, Japan). The immunohistochemical parameters assessed in each area detected included integrated optical density, total stained area, and mean optical density (MOD). Defined as MOD = integrated optical density/total stained area, MOD was taken to be equivalent to the intensity of stain in the positive cells. The immunoreactivity level was based on the MOD values.
The rate of cell proliferation was calculated as the percentage of anti-Ki-67-stained cells in relation to the total amount of stroma, glands, and blood vessels.
Apoptosis detection system
Apoptosis was quantified through terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling by using the In Situ Cell Death Detection Kit, POD (Roche Applied Science, Mannheim, Germany). Sections were treated according to the manufacturer’s protocol. Briefly, sections were deparaffinized and rehydrated, microwaved in citrate buffer (pH 6.0) for antigen retrieval, and treated with terminal deoxynucleotidyl transferase incubation buffer at 37°C for 60 minutes. As a negative control, a number of slides were subjected to treatment without terminal deoxynucleotidyl transferase. After the sections were treated with peroxidase-labeled antidigoxigenin antibody, the reaction was developed with 3,3′-diaminobenzidine. The percentage of apoptotic cells was evaluated by counting labeled cells at ×400 magnification in 30 randomly selected and homogeneous fields.
After the mice were sacrificed, a routine selection of tissues from major organ systems including the heart, liver, spleen, lung, and kidney were processed for histopathological examination. All selected tissues were stained with hematoxylin-eosin, and morphological changes of the tissues were identified by microscopy.
Statistical analyses were performed with the software (SPSS 13.0; IBM Corp, Armonk, NY). Data are expressed as mean ± SEM. All the results were analyzed by 1-way analysis of variance with post hoc tests for multiple comparisons and Student t test for paired comparisons. A P value < .05 was regard as statistically significant.
Establishment and verification of endometriosis model
The effectiveness of our mouse xenograft model of human endometriosis was analyzed by detection of the fluorescent reporter RFP or macroscopic observation. Eight days after the implantation of endometrial tissue, mice underwent fluorescence imaging or laparotomy and endometriotic lesions were found in 96% of animals (24/25). First, using the NightOWL LB 983 imaging system (Berthold Technologies), we found RFP-expressing endometriotic lesions in 48% of the mice that received the human tissue graft ( Figure 1 ). When the intraperitoneal cavity of these mice were analyzed, only the lesions in the cavity were fluorescent, with no background from non-RFP-expressing recipient tissue. The establishment of the disease in mice without detectable RFP expression was confirmed by intraperitoneal lesion formation. One animal was expelled from the experiment because there were no lesions observed when it was tracked at subsequent laparotomy. In mice in which lesions had developed, the lesions were highly vascularized as observed on gross morphological examination ( Figure 2 , A and B). The average number of lesions per mouse was 2 and ranged from 1-5 per mouse. Studies from our group as well as others revealed that the “take rate” and the number of lesions per animal is highly variable. The effectiveness of this xenograft model was also verified by histological examination of the lesions. Histologically, the murine lesions demonstrated the presence of human endometrial glands and stroma along with epithelial cells lining the lumen ( Figure 2 , C).
Inhibition of endometriosis growth by ENMD-1068
To evaluate the effect of ENMD-1068 treatment on endometriotic lesions, animals with surgically induced endometriosis that had been allowed to establish for 10 days were treated with 1 of 2 different doses of ENMD-1068 or with saline solution as a negative control. The size of established lesions in the high-dose (50 mg/kg) ENMD-1068 group was significantly smaller than that in the saline group (2.53 ± 0.61 mm 3 vs 13.87 ± 2.45 mm 3 , P < .01) ( Figure 2 , D). Moreover, the experiment showed that the effect of ENMD-1068 was dose dependent, because a lower dose (25 mg/kg) resulted in a smaller reduction in the volume of observed lesions (5.71 ± 0.93 mm 3 , P < .05) ( Figure 2 , D).
Effect of ENMD-1068 on IL-6 and MCP-1 levels
We next analyzed the effect of ENMD-1068 treatment on the release of IL-6 and MCP-1 from ectopic human endometrium harvested from the endometriosis model mice. The concentrations of both IL-6 and MCP-1 were remarkably increased in the saline-treated control group, while the increases were suppressed in both ENMD-1068 25 mg/kg ( P < .01 for IL-6; P < .05 for MCP-1) and ENMD-1068 50 mg/kg ( P < .01 for IL-6; P < .05 for MCP-1) treatment groups ( Figure 3 ). Moreover, the effect of ENMD-1068 on IL-6 expression was dose dependent. The IL-6 expression levels were significantly lower in the ENMD-1068 50 mg/kg group than in the ENMD-1068 25 mg/kg group ( P < .05) ( Figure 3 ).
Effect of ENMD-1068 on NF-κB and VEGF levels
To assess whether ENMD-1068 is effective in blocking the NF-κB pathway in vivo, NF-κB (RelA) immunohistochemistry was performed with all endometriotic lesions. RelA immunostaining was detected in the cytoplasm and the nucleus of the cells ( Figure 4 , A–C). Compared with control vehicle treatment, treatment with ENMD-1068 25 mg/kg and ENMD-1068 50 mg/kg for 5 days significantly decreased the expression of RelA ( P < .05) ( Figure 5 ). Furthermore, RelA expression in endometriotic lesions was significantly lower with ENMD-1068 50 mg/kg treatment than with ENMD-1068 25 mg/kg treatment ( P < .05) ( Figure 5 ).