Endometriosis is an estrogen-dependent inflammatory disease that affects 5% to 10% of reproductive-age women in the United States.11 Histologically, the defining characteristic of endometriosis is the ectopic presence of tissue (lesions) mimicking the appearance of uterine (eutopic) endometrium. These lesions are defined by endometrial-like glands, stroma, and hemosiderin, with blood vessels, nerve fibers, muscle, and immune cells.11 Retrograde menstruation is hypothesized to be the primary mechanism by which lesions are induced.32,36 During retrograde menstruation, damage-associated molecular patterns (DAMPs) can be released. Damage-associated molecular patterns activate resident immune cells, including macrophages.26,58 Activated immune cells produce proinflammatory mediators in an NF-κB–dependent manner, such as TNF-α, IL-1β, and IL-33 that are believed to contribute to pain.26,58 In response to these stimuli, nociceptor neurons further increase and maintain inflammation by secreting neuropeptides such as substance P and calcitonin gene–related peptide (CGRP),36,58 as well as chemokines such as CCL2.23 These mediators may also contribute to the recruitment of more immune cells and lesion growth.23,36
One key clinical feature of endometriosis is different forms of debilitating pain, including chronic pelvic pain, dysmenorrhea, dyspareunia, and dyschezia.32,36 Of these, pelvic pain represents a major clinical problem since women with that symptom report lower quality of life and mental health.14 Pain causes affected women to lose, on average, 10.8 hours of work weekly, primarily as a result of reduced effectiveness during working time.39 Pelvic organ cross-sensitization contributes, moreover, to diverse abdominal pain in disorders of the lower gut, pelvic urinary, or gynecologic organs leading to significant problems in diagnosing and treating these diseases.9
Most existing rodent models of endometriosis use surgery to either implant tissue or remove ovaries (ovariectomy).44 Ovaries are the main source of estrogen, and because endometriosis is an estrogen-dependent inflammatory disease, hormonal supplementation is required to maintain lesion growth.10,44,50 However, surgical incisions in the skin and deeper tissue increase guarding behavior in animals, indicating pain29,33 and can also increase peripheral and central neuronal spontaneous activity, suggesting these procedures influence animal behavior.59,60 Unfortunately, no model completely resembles all aspects of endometriosis, and pain is only rarely measured.44 When pain measurements are made, these rarely reflect the human experience. Usually, the primary focus is on evaluating evoked responses.10,44,50 An evoked response to thermal hyperalgesia is the most frequently used behavioral test in rodent models of endometriosis,10,44,50 but sensitivity to heat stimuli is rarely reported by women with endometriosis. The development of critically needed new endometriosis therapies is most likely to be successful if disease models recapitulate the human disease phenotype more accurately.
We have established a novel murine model for endometriosis-associated pain that induces neuronal and behavioral changes consistent with the disease phenotype in women. We validated this model by showing that both evoked and spontaneous pain measures were reduced by letrozole (aromatase inhibitor) and danazol (androgen), demonstrating that this model responds to drugs active in human disease.
2. Material and methods
2.1. Study design
Our objective was to standardize a nonsurgical mouse model of endometriosis-associated pain. Our strategy was to interrogate both spontaneous and evoked pain and to perform behavioral analysis of freely moving mice. Block randomization was used to randomize subjects into groups resulting in equal sample sizes.
Mice were treated daily by oral gavage with letrozole 5 mg/kg (cat #PHR1540; MilliporeSigma, Burlington, MA), danazol 35 mg/kg (cat #D8399; MilliporeSigma), or vehicle (5% Tween 80, 5% dimethyl sulfoxide [DMSO] in phosphate-buffered saline [PBS]) starting at day 29 and ending on day 56. All doses were defined based on previous studies showing efficacy at reducing endometriotic lesions45,64 All treatments were performed after behavioral testing. The investigators were blinded to the treatment groups in all testings until the end of the experiment and analysis.
All procedures were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee at Boston Children's Hospital (protocol number 16-12-3265). Healthy female C57BL/6 mice (Stock #000664, 8 weeks of age, 20 ± 2 g), purchased from Jackson Laboratories (Bar Harbor, ME), were used in this study. Mice were randomly assigned and housed in standard clear plastic cages with no more than 5 mice per cage in a 12:12-hour light/dark cycle with ad libitum water and food. Behavioral testing was performed between 9 am and 5 pm in a room maintained at a temperature of 21 ± 1°C. All efforts were made to minimize the number of animals used and their suffering. Euthanasia was performed by CO2 inhalation.
2.3. Induction of endometriosis
After at least 1 week of acclimatization, donor mice received a subcutaneous injection of 3-μg/mouse estradiol benzoate to stimulate the growth of the endometrium. Seven days later, the uteri of the donor mice were dissected into a Petri dish containing Hank's balanced salt solution (HBSS; Thermo Fisher Scientific, Waltham, MA) and split longitudinally with a pair of scissors. Uterine horns from each donor mouse were minced with scissors and scalpel one at the time, ensuring that the maximal diameter of each fragment was consistently smaller than 1 mm (Supplementary Fig. 1, available at http://links.lww.com/PAIN/A965). Each dissociated uterine horn was then injected intraperitoneally using an 18G needle (cat #305185 Thin wall; BD, Franklin Lakes, NJ) into a recipient mouse in 500 µL of HBSS. One donor mouse was used for every 2 endometriosis mice. Sham mice received an intraperitoneal injection of 500 µL of HBSS.
2.4. Histological analysis
Lesions were dissected at 14, 28, 42, and 56 days post-implantation (dpi) for histopathological analysis using H&E staining. Paraffin-embedded lesions were cut into 5-µm sections, and images were taken in an inverted light microscope using 20 and 40x objectives (Zeiss Axio Observer Z1; Carl Zeiss Microscopy, Thornwood, NY). For immunofluorescence, lesions or paired thoracic and lumbosacral (T10-L3 plus L6-S1) dorsal root ganglia (DRG) were dissected at 56 dpi and maintained in 4% paraformaldehyde (PFA, for 24 hours) and then to 30% sucrose (for 72 hours). These ganglia were chosen because of their involvement in abdominal pain and pelvic organ cross-sensitization.12,13,35 Optimum cutting temperature reagent (Tissue-Plus O.C.T.; Fisher Healthcare, Thermo Fisher Scientific)-embedded lesions were cut into 16- or 40-µM (only those involving PGP9.5 staining) sections. O.C.T.-embedded DRG were cut into 12-µm sections. Images were taken and processed on a confocal microscope using 20x objective (Zeiss LSM 880 laser scanning microscope with Airyscan; Carl Zeiss Microscopy). Primary antibodies used in this study are as follows: anti-CD31 (1:100, clone MEC 7.46, cat #ab7388; Abcam, Cambridge, MA), anti-ERα (1:500, cat #ab3575, Abcam), anti-PGP9.5 (1:50, cat #ab8189, Abcam), anti-CGRP (1:500, cat #C8198, MilliporeSigma), anti-TRPA1 (1:250, cat #ACC-037; Alomone Labs, Jerusalem, Israel), anti-TRPV1 (1:250, cat #ACC-030; Alomone Labs), anti-CD68 (1:500, cat #ab125212; Abcam), anti-GR1 (1:200, clone RB6-8C5; BioLegend, San Diego, CA), anti-Tryptase (1:100, cat #ab2378; Abcam), anti-CD45 (1:500, cat # ab10558; Abcam), and pNF-κB p65 (1:50, cat #sc-166748; Santa Cruz Biotechnology, Dallas, TX). Secondary antibodies used in this study were as follows: goat anti-mouse Alexa 488 (1:4000, cat #A-10680; Life Technologies, Carlsbad, CA, Thermo Fisher Scientific), goat anti-rat 647 (1:500, cat #A-21247; Life Technologies, Thermo Fisher Scientific), goat anti-rabbit DyLight 488 (1:200,cat #DI-1488, Vector Labs, Burlingame, CA), goat anti-rabbit DyLight 594 (1:200, cat #DI-1594 Vector Labs), and goat anti-mouse Alexa 488 (1:2000, cat #A-11001; Life Technologies, Thermo Fisher Scientific).
2.5. Calcium imaging
Calcium imaging of DRG neurons (T10-L3, L6-S1) was performed as previously described.41 Dorsal root ganglia were dissected into Neurobasal-A medium (Life Technologies, Thermo Fisher Scientific), dissociated in collagenase A (1 mg/mL)/dispase II (2.4 U/mL) (RocheApplied Sciences, Indianapolis, IN) in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-buffered saline (MilliporeSigma) for 70 minutes at 37°C. After trituration with glass Pasteur pipettes of decreasing size, DRG cells were centrifuged over a 10% bovine serum albumin (BSA) gradient, plated on laminin-coated cell culture dishes. Dorsal root ganglia were loaded with 5 μM of Fluo-4AM in Neurobasal-A medium, incubated for 30 minutes 37°C, washed with HBSS, and imaged in using an Eclipse Ti-S/L100 inverted microscope (Nikon Instruments, Melville, NY). An ultraviolet light source (Lambda XL lamp; Sutter Instrument, Novato, CA) was used for excitation of Fura-2-AM by alternating 340- and 380-nm wavelengths. NIS-elements software (Nikon Instruments) was used to image, process, and analyze 340/380 ratiometric images from DRG neurons. An increase in the 340/380 ratio of 10% or more from baseline levels was considered a positive response to a ligand. To assess TRPV1 or TRPA1 activation, DRG plates were recorded for 8 minutes, which was divided into 2 minutes of initial reading (0-second mark, baseline values), following at the 120-second mark by stimulation with capsaicin (500 nM, TRPV1 agonist) or allyl isothiocyanate (AITC) (100 µM, TRPA1 agonist) for 4 minutes, and finally, beginning at the 360-second mark, 2 minutes of depolarization with KCl (40 mM) to activate all neurons.
2.6. Western blot
Lesions or paired thoracic and lumbosacral (T10-L3 plus L6-S1) DRG neurons from sham and endo mice were dissected at 56 dpi. Protein levels were quantified using Pierce BCA Protein Assay Kit (Life Technologies, Thermo Fisher Scientific). Western blot was performed using 15 µg of protein blotted onto polyvinylidene fluoride (PVDF) membranes. After blocking with 5% dry milk in tris-buffered saline-Tween (TBS-T) (Tween 0.01%), pH 7.4, we probed with the following primary antibodies in blocking buffer: anti-TRPA1 (1:2000, cat #ACC-037; Alomone Labs), anti-TRPV1 (1:2000, cat #ACC-030; Alomone Labs), and anti-β-actin (1:4000, cat #A3854; MilliporeSigma). After washing with TBST, blots were incubated with secondary antibody in blocking buffer. Secondary antibodies used were as follows: goat anti-rabbit (1:8000, cat #1706515; Bio-Rad, Hercules, CA) and goat anti-mouse (1:8000, cat #1706516; Bio-Rad). Image J software (NIH, Bethesda, MD) was used to measure the optical density of the bands.
2.7. Behavioral testing
For mechanical and heat hyperalgesia tests, mice were allowed to habituate to the apparatus for at least 2 hours and during 3 consecutive days before the beginning of measurements. After habituation, baseline measurements were obtained on 2 consecutive days before the induction of endometriosis. Pain intensity to a mechanical stimulus (mechanical hyperalgesia) in the abdominal region was measured using von Frey filaments. The experimenter was trained, care was taken not to stimulate the same point consecutively, and the stimulation of the external genitalia was avoided. A jump or paw flinch was considered a withdrawal response.21,30 The mechanical threshold was determined by the up and down method starting with 0.4-g filament and calculated using the open-source software Up-Down Reader.20
To measure pain sensitivity to a heat stimulus (heat hyperalgesia), mice were placed on the temperature-controlled (29°C) glass plate of a Hargreaves apparatus (Model 390G; IITC Life Science, Woodland Hills, CA). A radiant heat source was used to stimulate the paw by gradually increasing the temperature of the plantar surface. The threshold of pain was determined as the latency (in seconds) to evoke a response of paw withdrawal: paw flinches or licking. An exposure limit of 15 seconds was used to prevent tissue damage.
For the dynamic weight-bearing assay, mice were allowed to habituate to the apparatus for 15 minutes and during 3 consecutive days before the beginning of measurements. Mice were gently placed in a small Plexiglas chamber (11.0 × 19.7 × 11.0 cm) with floor sensors containing pressure transducers (Model BIO-DWB-M; Bioseb, Vitrolles, France). The system uses software that records the average weight that each paw exerts on the floor (in grams), without any interference of the investigator. For the testing, the mouse was placed in the chamber and allowed to move freely within the apparatus for a period of 5 minutes.
For the thermal gradient assay, a continuous temperature gradient (7-50°C) was established along with a metallic base plate where the mice were gently placed. Mice then walked freely while being video-recorded from above (Bioseb) as described previously.1 After an exploration period (30 minutes), individual mice showed a distinct preference, indicating the most comfortable temperature range. Data are presented by the time spent (in seconds) on each zone set at specific temperatures during 60 minutes, and the absolute sum of the squared difference from the mean (ie, sample variance) was used to determine data dispersion. For each treatment, the best-fitted Gaussian curve was also applied to have a better readout of data dispersion. Both the absolute sum of squares and best-fitted Gaussian curve were based on the last 30 minutes of the thermal gradient assay as described previously.1 Each run lasted 1.5 hours, and 2 mice were simultaneously recorded in separate corridors.
For spontaneous abdominal pain measurements, licking of the abdomen, stretching the abdomen (abdominal contortions), and squashing of the lower abdomen against the floor were quantified as previously described.16,21,30 Briefly, direct abdominal licking was quantified during 10 minutes using bottom-up video recording. We counted the number of times the mouse directly groomed the abdominal region without going for any other region before or after the behavior. For abdominal contortions, mice were placed in individual chambers in a temperature-controlled (29°C) glass plate, and the number of abdominal contortions was quantified for 10 minutes. Positive responses consist of a contraction of the abdominal muscle together with stretching of hind limbs. For abdominal squashing, the number of times the mice pressed the lower abdominal region against the floor in 5 minutes was quantified. All behaviors were confirmed in a dark chamber using bottom-up video without the presence of an investigator. In all testing, the investigators were blinded to the treatments.
2.8. Cytokine measurements
Lesions were dissected at 56 dpi and homogenized in 500 µL of the appropriate buffer containing protease inhibitors. Because sham mice do not develop lesions, uterine horns from donor mice were used to determine baseline levels of each cytokine. TNF-α, IL-1β, and IL-33 levels were determined by enzyme-linked immunosorbent assay (ELISA) kits (Thermo Fisher Scientific) accordingly with manufacturer instructions. Protein levels were quantified using Pierce BCA Protein Assay Kit (Life Technologies, Thermo Fisher Scientific). The results are expressed as picograms (pg) of each cytokine per mg of protein.
2.9. Lesion size
Lesions were carefully dissected at 56 dpi and measured using calipers. Lesion size is expressed in millimeters calculated from the mean of 2 measurements (width and height).
2.10. Statistical analysis
Results are presented as mean ± SEM. Data were analyzed using the software GraphPad Prism version 6.01 (GraphPad Software, San Diego, CA). Two-way repeated-measure analysis of variance, followed by Tukey's post hoc, was used to analyze data from experiments of multiple time points (dynamic weight bearing, mechanical, and thermal hyperalgesia). One-way analysis of variance followed by Tukey's post hoc was used to analyze data from experiments with a single time point. Comparison between 2 groups was conducted using the Student t-test. For the percentage of mice with visible lesions, statistical analysis was estimated by the Kaplan–Meier method followed by the log-rank test. For all analyses, statistical differences were considered significant when P < 0.05.
3.1. Mouse model of endometriosis-associated pain induces lesions that resemble human lesions
To induce endometriosis-like lesions without surgery, we dissociated the horn of a donor mouse and injected the resulting cells and tissue fragments into the peritoneal space of a recipient animal. We then performed histological analysis using H&E staining to determine whether this model leads to the formation of lesions that resemble those found in women. We observed throughout the course of this model, the presence of lesions with endometrial-like glands, stromal cells, and blood vessels (Fig. 1A). Fifty-six dpi, we observed visible lesions in 90% of the mice (Fig. 1B). These lesions varied in shape, color, size, and location, which correlates with the variation observed in women (Fig. 1C). Because endometriosis is an estrogen-dependent disease,11 we evaluated estrogen receptor (ER) expression in lesions, and by immunofluorescence, we observed the expression of ERα 14 to 56 dpi. Given the role of cytokines in both endometriosis and pain,36,53 we next evaluated the levels of TNF-α, IL-1β, and IL-33. These cytokines were chosen because they are increased in the lesions of women with endometriosis,36 nociceptor neurons express their receptors,6,27,34 and they are directly involved in pain signaling in other models.25,54,57,63 Because sham mice do not develop lesions (Fig. 1B), uterine horns from donor mice were used to measure baseline levels for each cytokine. We observed an increase in TNF-α, IL1-β, and IL-33 levels in the ectopic lesions when compared with the uterine horn from donor mice (Fig. 1D). Finally, given the key role of angiogenesis in lesion growth and maintenance, blood vessel presence was confirmed by anti-CD31 staining (Fig. 1E). Thus, the lesions observed in our model recapitulate key features found in women with endometriosis.
3.2. Mouse lesions show the presence of immune cells close to nerve fibers
Pain is the most debilitating symptom of endometriosis. In endometriotic lesions, the presence of CGRP-, TRPA1-, and TRPV1-expressing fibers8,22,37,51 and the presence of immune cells, such as macrophages, neutrophils, and mast cells,2,26,36 are key factors for pain generation. Therefore, we used immunofluorescence to investigate whether nerve fibers and immune cells are present in the lesions. Lesions were double stained using PGP9.5 (neuronal pan marker) together with CGRP, TRPA1, or TRPV1. We found CGRP-, TRPA1-, and TRPV1-expressing nerve fibers in the lesions (Fig. 2A). We next addressed whether immune cells are positioned near the nerve fibers. Costaining with CD45 (pan-immune cell marker) and PGP9.5 demonstrated that CD45-positive cells were in proximity to PGP9.5-positive nerve fibers (Fig. 2B). Additional staining showed the presence of macrophages (CD68), mast cells (tryptase), and neutrophils (GR1) (Fig. 2C). These results demonstrate that the nerve fibers and immune cells, which are a common finding in human lesions, are also found in our mouse endometriotic lesions.
3.3. Endometriosis induces NF-κB activation and increases responsiveness in TRPA1- and TRPV1-expressing dorsal root ganglion neurons
Chronic inflammation induces changes in the neuronal pain circuit leading to peripheral sensitization, which is defined by a reduction in the activation threshold and/or an increase in the magnitude of responsiveness at the peripheral ends of nociceptor neurons.18,42 To determine whether endometriosis produces neuronal changes consistent with peripheral sensitization, we performed calcium imaging using dorsal root ganglion (DRG) neurons dissected from mice with endometriosis and shams. Activation of DRG neurons can be observed through an increase in calcium influx upon stimulation.17,41 Because we observed the presence of TRPA1- and TRPV1-expressing nerve fibers in the lesions, we investigated the activation of these 2 subpopulations of DRG neurons by using AITC (TRPA1 agonist) or capsaicin (Caps, TRPV1 agonist). Both AITC (Figs. 3A–D) and capsaicin (Figs. 3E–H) increased the percentage of responding neurons (Figs. 3A and E) and the amplitude of the response (Figs. 3B, C and F, G) of DRG neurons from mice with endometriosis compared with sham-treated animals. These changes indicate that endometriosis modified the responsiveness of TRPA1- and TRPV1-expressing DRG neurons. Extending these results, we observed increased TRPA1 and TRPV1 protein levels (Figs. 3D and H) and colocalization with phosphorylated NF-κB p65 subunit (Figs. 3I and J). We conclude that, in our model, TRPA1- and TRPV1-expressing DRG neurons increase their responses to chemical irritants, likely as a result of increased expression of cognate ion channels.
3.4. Endometriosis-associated pain is sensitive to treatment with drugs known to have clinical activity
We next determined whether experimental induction of endometriosis lesions induces pain-related behavioral changes in mice. We first assessed this using von Frey filaments (Fig. 4A) looking for mechanical sensitivity of the abdominal wall, dynamic weight bearing (Fig. 4B), and the Hargreaves assay (thermal sensitivity of the paws) (Fig. 4C). These techniques are capable of detecting abdominal pain in different experimental models.21,30,31,61 We observed an increase in evoked pain only using von Frey filaments (Fig. 4A), as observed by a decrease in the abdominal mechanical threshold. We did not observe any postural change using dynamic weight bearing (Fig. 4B) or any change in thermal hyperalgesia using the Hargreaves assay (Fig. 4C). Moreover, we did not observe any change on mouse weight throughout the course of the model (Fig. 4D). Based on that result, the von Frey test was the only evoked-pain method used going forward.
Next, we investigated whether our model could predict the efficacy of clinically active drugs, such as letrozole and danazol.46,48 We observed that both letrozole and danazol reduced endometriosis-associated pain (Fig. 5A). Letrozole and danazol also reduced the percentage of mice with visible lesions when compared with the vehicle-treated group (Fig. 5B). However, the remaining lesions were not significantly smaller than control lesions (Fig. 5C). These results show that estrogen-reducing drugs such as letrozole (an aromatase inhibitor) and danazol (an androgen) reduce lesion number and ameliorate endometriosis-associated evoked pain.
3.5. Letrozole and danazol normalizes endometriosis-associated changes on thermal selection
We next used a thermal gradient assay to measure the mouse's own determination of general discomfort.1,52 This assay uses video recordings to measure behavior in freely moving mice without the presence of a human investigator.1 Sham mice prefer temperatures 27 to 36°C with a higher preference for 34°C, whereas mice with endometriosis exhibited a more dispersed pattern from 21 to 36°C with no single preferred temperature (Figs. 6A–C). To quantify the data dispersion, variance in location (temperature) was calculated (Fig. 6B). Mice with endometriosis presented a higher variance, and this change in normal thermal selection was alleviated by letrozole and danazol with a return to a strong preference for 34 to 36°C.
3.6. Endometriosis induces spontaneous pain
Clinical pain, in contrast to acute nociceptive pain, often is accompanied by ongoing or intermittent spontaneous pain.15 However, existing models of endometriosis have not been shown to induce spontaneous pain.10,44,50 Abdominal contortions induced by chemicals (eg, acetic acid) have been widely used for modeling visceral pain7,49 and we wondered whether the disease-based model of endometriosis generates spontaneous abdominal pain. We found that mice with endometriosis lesions exhibit spontaneous pain-like behaviors such as direct abdominal licking (Fig. 7A), abdominal squashing (Fig. 7B, pressuring of the lower abdominal region against the floor), and abdominal contortions (Fig. 7C, contraction of the abdominal muscle together with stretching of hind limbs) at all evaluated time points. Both letrozole and danazol decreased these behaviors, demonstrating that this model of endometriosis induces spontaneous pain and that this pain can be reduced by clinically active drugs.
Although pain is the major symptom of endometriosis in most patients, investigators developing existing nonsurgical models of endometriosis have not reported any attempt to measure spontaneous pain.10,44,50 Indeed, only a minority of studies measure any indication of pain at all. Stimulus-evoked responses at best correlate only with a few aspects of pain, such as withdrawal from a stimulus or learned behavioral avoidance from a potentially painful situation.15 This evoked type of response relates to only aspects of the symptoms reported by chronic pain patients, who more often report spontaneous ongoing or intermittent pain,15,56 which the brain processes differently vs evoked pain.40 Moreover, most drugs showing analgesic effects in preclinical models that only measure evoked pain have failed in clinical trials3,56,62; this includes drugs targeting endometriosis.4 These results together indicate that although preclinical models evaluating evoked pain may have some utility in the discovery of novel targets or mechanisms of action of new drugs, their ability to predict efficacy in humans is low. There is an unmet need for a disease model of chronic spontaneous pain.
Here, we describe a nonsurgical mouse model of endometriosis that recapitulates many features of the disease phenotype observed in women, especially regarding spontaneous pain. Importantly, both disease and pain phenotypes in the model respond to clinically active drugs. The ability to control the composition and timing of the inoculum may enable studies that contribute to understanding mechanisms of disease development as well as the identification of clinically useful prophylactic treatments. We selected C57BL/6 mice for the model to bring the plethora of genome-manipulated strains available in this background to bear on understanding the pathophysiology of endometriosis-associated pain thereby identifying new drug targets. This includes the use of transgenic mice expressing fluorescence and/or luminescent proteins as tissue donors to evaluate and locate lesions in live recipient mice. Because of their small size and large litters, mice are also cost effective in comparison with other animals. Finally, because this model generates both evoked and spontaneous pain, it is also likely to be useful in predicting the activity of novel drugs to treat existing endometriosis.
We observed an increase in the levels of TNF-α, IL-1β, and IL-33 in the lesion tissue. The presence of cytokines and immune cells contributes to pain generation in endometriosis; TNF-α, IL-1β, and IL-33 are increased in women with endometriosis36 and increases in either TNF-α or IL-1β both correlate with chronic pelvic pain.36 These cytokines can act directly on nociceptor neurons6,27,34 thereby initiating pain. In addition, these cytokines act as chemoattractants for immune cells. For instance, in rheumatoid arthritis, TNF-α increases the expression of ST2 (IL-33 receptor) favoring IL-33–induced neutrophil recruitment toward inflammatory foci.55 IL-33 also stimulates proliferation and vascularization of surgically induced endometriosis lesions in mice.38 Once in the lesion, mast cells, neutrophils, and macrophages produce additional inflammatory mediators that contribute to the maintenance and growth of the lesion.2,26,36 Indeed, reduction of neutrophil recruitment to endometriosis lesions in formyl peptide receptor 1 knockout mice decreases evoked pain, lesion size, mast cell activation, and IL-1β production.19 The increased cytokine concentrations we observed likely contribute, therefore, to both lesion maintenance and pain generation.
The presence of nerve fibers in lesions is another hallmark of endometriosis.8,22,37,51 A growing body of evidence demonstrates that immune cells and nociceptor interaction are important for the development of acute and chronic inflammatory diseases,42 including endometriosis.5,23,36,58 For instance, estradiol stimulates the release of chemokines such as CCL2 by nociceptors. This recruits macrophages toward nerve fibers.23 Here, we demonstrate that immune cells are in close proximity to nerve fibers, which likely contribute to pain both in our model and in human disease. Specifically, the main nerve fibers found in women's endometriotic lesions are CGRP-, TRPA1-, and TRPV1-positive fibers,8,22,37,51 and increased expression of TRPV1 in human lesions correlates with chronic pain.43 Corroborating these clinical data, we show that nerve fibers in mouse endometriotic lesions express both TRPA1 and TRPV1 channels.8,43 Furthermore, we observed a pattern of neuronal activation consistent with peripheral sensitization18,42 as detected by an increase in NF-κB activation28,47 and calcium imaging in DRG neurons.17,41
We interrogated pain or general discomfort in our model using both evoked and spontaneous pain behaviors and also assays allowing behavioral choice in freely moving mice (thermal gradient assay).1 Among the spontaneous pain behaviors, we detected direct abdominal licking, abdominal squashing, and abdominal contortions. The latter has been widely used for the discovery of novel drugs targeting visceral pain.7,24 However, experimental disease models that cause abdominal pain (eg, colitis, pancreatitis, etc) rarely report abdominal contortions or squashing. Instead, injection of chemical stimuli such as AITC, capsaicin, acetic acid, phenyl-p-benzoquinone, or other chemicals is required to induce spontaneous abdominal pain in mice.49 Such chemical injections are generally acute and do not adequately reproduce the chronic pain observed in most patients with abdominal pain, which is generally caused by chronic disease.49 We find that endometriosis induced direct abdominal licking, contortions, and squashing, and these behaviors were reduced by letrozole and danazol. Thus, because it is among the first disease models of spontaneous pain, the utility of our model may extend beyond endometriosis to abdominal and visceral pain more generally.
We have established a model of endometriosis-associated pain that responds to clinically active drugs, such as letrozole (aromatase inhibitor) and danazol (androgen). Importantly, this model recapitulates key features of the disease phenotype in women, such as the presence of endometriotic-like lesions with nerve fibers, blood vessels, immune cells, and proinflammatory cytokines. Interrogating spontaneous pain and evoked pain in this model, together with behavioral analysis of freely moving mice, should help facilitate the development of novel therapeutic approaches for endometriosis, including analgesic drugs effective for chronic pain.
Conflict of interest statement
The authors have no conflicts of interest to declare.
Appendix A. Supplemental digital content
Supplemental digital content associated with this article can be found online at http://links.lww.com/PAIN/A965.
Supplemental video content
A video abstract associated with this article can be found at http://links.lww.com/PAIN/A964.
V. Fattori thanks the National Council for Scientific and Technological Development (CNPq, Brazil) for the 12-month scholarship to develop his Split Fellowship (Doutorado Sanduiche SWE, CNPq). W.A. Verri also acknowledges CNPq Researcher fellowship. The authors are grateful for the technical assistance of the Boston Children's Hospital Neurodevelopmental Behavioral Core (NBC) on the thermal gradient assay. This work was supported by grants from The J. Willard and Alice S. Marriott Foundation.
Author contributions: conceptualization: V. Fattori and MSR; investigation and data analysis: V. Fattori, N.S. Franklin, R. Gonzalez-Cano, D. Peterse, A. Ghalali, E. Madrian, and N. Andrews; methodology: V. Fattori, W.A. Verri, C. J. Woolf, and M. S. Rogers; funding acquisition: M. S. Rogers; resources: C. J. Woolf and M. S. Rogers; software: R. Gonzalez-Cano; project administration: V. Fattori; supervision: V. Fattori and M. S. Rogers; writing—original draft: V. Fattori; writing—reviewing and editing manuscript: all authors. All authors read and approved the final version of the manuscript.
. Alexandre C, Latremoliere A, Ferreira A, Miracca G, Yamamoto M, Scammell TE, Woolf CJ. Decreased alertness due to sleep loss increases pain sensitivity in mice. Nat Med 2017;23:768–74.
. Anaf V, Chapron C, El Nakadi I, De Moor V, Simonart T, Noel JC. Pain, mast cells, and nerves in peritoneal, ovarian, and deep infiltrating endometriosis. Fertil Steril 2006;86:1336–43.
. Andrews NA, Latremoliere A, Basbaum AI, Mogil JS, Porreca F, Rice AS, Woolf CJ, Currie GL, Dworkin RH, Eisenach JC, Evans S, Gewandter JS, Gover TD, Handwerker H, Huang W, Iyengar S, Jensen MP, Kennedy JD, Lee N, Levine J, Lidster K, Machin I, McDermott MP, McMahon SB, Price TJ, Ross SE, Scherrer G, Seal RP, Sena ES, Silva E, Stone L, Svensson CI, Turk DC, Whiteside G. Ensuring transparency and minimization of methodologic bias in preclinical pain research: PPRECISE considerations. PAIN 2016;157:901–9.
. Becker CM, Gattrell WT, Gude K, Singh SS. Reevaluating response and failure of medical treatment of endometriosis: a systematic review. Fertil Steril 2017;108:125–36.
. Binda MM, Donnez J, Dolmans MM. Targeting mast cells: a new way to treat endometriosis. Expert Opin Ther Targets 2017;21:67–75.
. Binshtok AM, Wang H, Zimmermann K, Amaya F, Vardeh D, Shi L, Brenner GJ, Ji RR, Bean BP, Woolf CJ, Samad TA. Nociceptors are interleukin-1beta sensors. J Neurosci 2008;28:14062–73.
. Blumberg H, Wolf PS, Dayton HB. Use of writhing test for evaluating analgesic activity of narcotic antagonists. Proc Soc Exp Biol Med 1965;118:763–6.
. Bohonyi N, Pohoczky K, Szalontai B, Perkecz A, Kovacs K, Kajtar B, Orban L, Varga T, Szegedi S, Bodis J, Helyes Z, Koppan M. Local upregulation of transient receptor potential ankyrin 1 and transient receptor potential vanilloid 1 ion channels in rectosigmoid deep infiltrating endometriosis. Mol Pain 2017;13:1744806917705564.
. Brumovsky PR, Gebhart GF. Visceral organ cross-sensitization - an integrated perspective. Auton Neurosci 2010;153:106–15.
. Bruner-Tran KL, Mokshagundam S, Herington JL, Ding T, Osteen KG. Rodent models of experimental endometriosis: identifying mechanisms of disease and therapeutic targets. Curr Womens Health Rev 2018;14:173–88.
. Bulun SE. Endometriosis. N Engl J Med 2009;360:268–79.
. Christianson JA, Liang R, Ustinova EE, Davis BM, Fraser MO, Pezzone MA. Convergence of bladder and colon sensory innervation occurs at the primary afferent level. PAIN 2007;128:235–43.
. Christianson JA, Traub RJ, Davis BM. Differences in spinal distribution and neurochemical phenotype of colonic afferents in mouse and rat. J Comp Neurol 2006;494:246–59.
. Facchin F, Barbara G, Saita E, Mosconi P, Roberto A, Fedele L, Vercellini P. Impact of endometriosis on quality of life and mental health: pelvic pain makes the difference. J Psychosom Obstet Gynaecol 2015;36:135–41.
. Fattori V, Borghi SM, Rossaneis AC, Bertozzi MM, Cunha TM, Verri WA Jr. Neuroimmune regulation of pain and inflammation: targeting glial cells and nociceptor sensory neurons interaction. In: Atta-ur-Rahman, Choudhary MI, editors. Frontiers in CNS drug discovery. Vol. 3: Bentham Science (UAE), 2017. p. 146–200.
. Fattori V, Pinho-Ribeiro FA, Borghi SM, Alves-Filho JC, Cunha TM, Cunha FQ, Casagrande R, Verri WA Jr. Curcumin inhibits superoxide anion-induced pain-like behavior and leukocyte recruitment by increasing Nrf2 expression and reducing NF-kappaB activation. Inflamm Res 2015;64:993–1003.
. Fattori V, Pinho-Ribeiro FA, Staurengo-Ferrari L, Borghi SM, Rossaneis AC, Casagrande R, Verri WA Jr. The specialized pro-resolving lipid mediator Maresin-1 reduces inflammatory pain with a long-lasting analgesic effect. Br J Pharmacol 2019;176:1728–44.
. Ferreira SH. Prostaglandins, aspirin-like drugs and analgesia. Nat New Biol 1972;240:200–3.
. Fusco R, D'Amico R, Cordaro M, Gugliandolo E, Siracusa R, Peritore AF, Crupi R, Impellizzeri D, Cuzzocrea S, Di Paola R. Absence of formyl peptide receptor 1 causes endometriotic lesion regression in a mouse model of surgically-induced endometriosis. Oncotarget 2018;9:31355–66.
. Gonzalez-Cano R, Boivin B, Bullock D, Cornelissen L, Andrews N, Costigan M. Up-down reader: an open source program for efficiently processing 50% von Frey thresholds. Front Pharmacol 2018;9:433.
. Gonzalez-Cano R, Merlos M, Baeyens JM, Cendan CM. sigma1 receptors are involved in the visceral pain
induced by intracolonic administration of capsaicin in mice. Anesthesiology 2013;118:691–700.
. Greaves E, Grieve K, Horne AW, Saunders PT. Elevated peritoneal expression and estrogen regulation of nociceptive ion channels in endometriosis. J Clin Endocrinol Metab 2014;99:E1738–1743.
. Greaves E, Temp J, Esnal-Zufiurre A, Mechsner S, Horne AW, Saunders PT. Estradiol is a critical mediator of macrophage-nerve cross talk in peritoneal endometriosis. Am J Pathol 2015;185:2286–97.
. Gregory NS, Harris AL, Robinson CR, Dougherty PM, Fuchs PN, Sluka KA. An overview of animal models of pain: disease models and outcome measures. J Pain 2013;14:1255–69.
. Inoue A, Ikoma K, Morioka N, Kumagai K, Hashimoto T, Hide I, Nakata Y. Interleukin-1beta induces substance P release from primary afferent neurons through the cyclooxygenase-2 system. J Neurochem 1999;73:2206–13.
. Izumi G, Koga K, Takamura M, Makabe T, Satake E, Takeuchi A, Taguchi A, Urata Y, Fujii T, Osuga Y. Involvement of immune cells in the pathogenesis of endometriosis. J Obstet Gynaecol Res 2018;44:191–8.
. Jin X, Gereau RW IV. Acute p38-mediated modulation of tetrodotoxin-resistant sodium channels in mouse sensory neurons by tumor necrosis factor-alpha. J Neurosci 2006;26:246–55.
. Kanngiesser M, Haussler A, Myrczek T, Kusener N, Lim HY, Geisslinger G, Niederberger E, Tegeder I. Inhibitor kappa B kinase beta dependent cytokine upregulation in nociceptive neurons contributes to nociceptive hypersensitivity after sciatic nerve injury. J Pain 2012;13:485–97.
. Klinger MB, Sacks S, Cervero F. A role for extracellular signal-regulated kinases 1 and 2 in the maintenance of persistent mechanical hyperalgesia in ovariectomized mice. Neuroscience 2011;172:483–93.
. Laird JM, Martinez-Caro L, Garcia-Nicas E, Cervero F. A new model of visceral pain
and referred hyperalgesia in the mouse. PAIN 2001;92:335–42.
. Laux-Biehlmann A, Boyken J, Dahllof H, Schmidt N, Zollner TM, Nagel J. Dynamic weight bearing as a non-reflexive method for the measurement of abdominal pain
in mice. Eur J Pain 2016;20:742–52.
. Laux-Biehlmann A, d'Hooghe T, Zollner TM. Menstruation pulls the trigger for inflammation and pain in endometriosis. Trends Pharmacol Sci 2015;36:270–6.
. Li LH, Wang ZC, Yu J, Zhang YQ. Ovariectomy results in variable changes in nociception, mood and depression in adult female rats. PLoS One 2014;9:e94312.
. Liu B, Tai Y, Achanta S, Kaelberer MM, Caceres AI, Shao X, Fang J, Jordt SE. IL-33/ST2 signaling excites sensory neurons and mediates itch response in a mouse model of poison ivy contact allergy. Proc Natl Acad Sci U S A 2016;113:E7572–9.
. Malykhina AP, Qin C, Greenwood-van Meerveld B, Foreman RD, Lupu F, Akbarali HI. Hyperexcitability of convergent colon and bladder dorsal root ganglion neurons after colonic inflammation: mechanism for pelvic organ cross-talk. Neurogastroenterol Motil 2006;18:936–48.
. McKinnon BD, Bertschi D, Bersinger NA, Mueller MD. Inflammation and nerve fiber interaction in endometriotic pain. Trends Endocrinol Metab 2015;26:1–10.
. Mechsner S, Schwarz J, Thode J, Loddenkemper C, Salomon DS, Ebert AD. Growth-associated protein 43-positive sensory nerve fibers accompanied by immature vessels are located in or near peritoneal endometriotic lesions. Fertil Steril 2007;88:581–7.
. Miller JE, Monsanto SP, Ahn SH, Khalaj K, Fazleabas AT, Young SL, Lessey BA, Koti M, Tayade C. Interleukin-33 modulates inflammation in endometriosis. Sci Rep 2017;7:17903.
. Nnoaham KE, Hummelshoj L, Webster P, d'Hooghe T, de Cicco Nardone F, de Cicco Nardone C, Jenkinson C, Kennedy SH, Zondervan KT; World Endometriosis Research Foundation Global Study of Women's Health
consortium. Impact of endometriosis on quality of life and work productivity: a multicenter study across ten countries. Fertil Steril 2011;96:366–73 e368.
. Parks EL, Geha PY, Baliki MN, Katz J, Schnitzer TJ, Apkarian AV. Brain activity for chronic knee osteoarthritis: dissociating evoked pain from spontaneous pain. Eur J Pain 2011;15:843.e841–814.
. Pinho-Ribeiro FA, Baddal B, Haarsma R, O'Seaghdha M, Yang NJ, Blake KJ, Portley M, Verri WA, Dale JB, Wessels MR, Chiu IM. Blocking neuronal signaling to immune cells treats streptococcal invasive infection. Cell 2018;173:1083–97.e1022.
. Pinho-Ribeiro FA, Verri WA Jr, Chiu IM. Nociceptor sensory neuron-immune interactions in pain and inflammation. Trends Immunol 2017;38:5–19.
. Rocha MG, e Silva JC, Ribeiro da Silva A, Candido Dos Reis FJ, Nogueira AA, Poli-Neto OB. TRPV1 expression on peritoneal endometriosis foci is associated with chronic pelvic pain. Reprod Sci 2011;18:511–15.
. Simitsidellis I, Gibson DA, Saunders PTK. Animal models of endometriosis: replicating the aetiology and symptoms of the human disorder. Best Pract Res Clin Endocrinol Metab 2018;32:257–69.
. Sinha S, Kaseta J, Santner SJ, Demers LM, Bremmer WJ, Santen RJ. Effect of CGS 20267 on ovarian aromatase and gonadotropin levels in the rat. Breast Cancer Res Treat 1998;48:45–51.
. Slopien R, Meczekalski B. Aromatase inhibitors in the treatment of endometriosis. Prz Menopauzalny 2016;15:43–7.
. Souza GR, Cunha TM, Silva RL, Lotufo CM, Verri WA Jr, Funez MI, Villarreal CF, Talbot J, Sousa LP, Parada CA, Cunha FQ, Ferreira SH. Involvement of nuclear factor kappa B in the maintenance of persistent inflammatory hypernociception. Pharmacol Biochem Behav 2015;134:49–56.
. Stratton P, Berkley KJ. Chronic pelvic pain and endometriosis: translational evidence of the relationship and implications. Hum Reprod Update 2011;17:327–46.
. Tappe-Theodor A, Kuner R. Studying ongoing and spontaneous pain in rodents—challenges and opportunities. Eur J Neurosci 2014;39:1881–90.
. Tirado-Gonzalez I, Barrientos G, Tariverdian N, Arck PC, Garcia MG, Klapp BF, Blois SM. Endometriosis research: animal models for the study of a complex disease. J Reprod Immunol 2010;86:141–7.
. Tokushige N, Markham R, Russell P, Fraser IS. Different types of small nerve fibers in eutopic endometrium and myometrium in women with endometriosis. Fertil Steril 2007;88:795–803.
. Touska F, Winter Z, Mueller A, Vlachova V, Larsen J, Zimmermann K. Comprehensive thermal preference phenotyping in mice using a novel automated circular gradient assay. Temperature (Austin) 2016;3:77–91.
. Verri WA Jr, Cunha TM, Parada CA, Poole S, Cunha FQ, Ferreira SH. Hypernociceptive role of cytokines and chemokines: targets for analgesic drug development? Pharmacol Ther 2006;112:116–38.
. Verri WA Jr, Cunha TM, Parada CA, Wei XQ, Ferreira SH, Liew FY, Cunha FQ. IL-15 mediates immune inflammatory hypernociception by triggering a sequential release of IFN-gamma, endothelin, and prostaglandin. Proc Natl Acad Sci U S A 2006;103:9721–5.
. Verri WA Jr, Souto FO, Vieira SM, Almeida SC, Fukada SY, Xu D, Alves-Filho JC, Cunha TM, Guerrero AT, Mattos-Guimaraes RB, Oliveira FR, Teixeira MM, Silva JS, McInnes IB, Ferreira SH, Louzada-Junior P, Liew FY, Cunha FQ. IL-33 induces neutrophil migration in rheumatoid arthritis and is a target of anti-TNF therapy. Ann Rheum Dis 2010;69:1697–703.
. Woolf CJ. Overcoming obstacles to developing new analgesics. Nat Med 2010;16:1241–7.
. Woolf CJ, Allchorne A, Safieh-Garabedian B, Poole S. Cytokines, nerve growth factor and inflammatory hyperalgesia: the contribution of tumour necrosis factor alpha. Br J Pharmacol 1997;121:417–24.
. Wu J, Xie H, Yao S, Liang Y. Macrophage and nerve interaction in endometriosis. J Neuroinflammation 2017;14:53.
. Xu J, Brennan TJ. Comparison of skin incision vs. skin plus deep tissue incision on ongoing pain and spontaneous activity in dorsal horn neurons. PAIN 2009;144:329–39.
. Xu J, Brennan TJ. Guarding pain and spontaneous activity of nociceptors after skin versus skin plus deep tissue incision. Anesthesiology 2010;112:153–64.
. Yan D, Liu X, Guo SW. The establishment of a mouse model of deep endometriosis. Hum Reprod 2019;34:235–47.
. Yekkirala AS, Roberson DP, Bean BP, Woolf CJ. Breaking barriers to novel analgesic drug development. Nat Rev Drug Discov 2017;16:810.
. Zarpelon AC, Rodrigues FC, Lopes AH, Souza GR, Carvalho TT, Pinto LG, Xu D, Ferreira SH, Alves-Filho JC, McInnes IB, Ryffel B, Quesniaux VF, Reverchon F, Mortaud S, Menuet A, Liew FY, Cunha FQ, Cunha TM, Verri WA Jr. Spinal cord oligodendrocyte-derived alarmin IL-33 mediates neuropathic pain. FASEB J 2016;30:54–65.
. Zhang X, Yuan H, Deng L, Hu F, Ma J, Lin J. Evaluation of the efficacy of a danazol-loaded intrauterine contraceptive device on adenomyosis in an ICR mouse model. Hum Reprod 2008;23:2024–30.