Elevated peritoneal expression and estrogen regulation of nociceptive ion channels in endometriosis

Context: Ovarian suppression is a common treatment for endometriosis-associated pelvic pain. Its exact mechanism of action is poorly understood, although it is assumed to reflect reduced pro-duction/action of estrogens. Objective: The objective of the study was to measure the expression of mRNAs encoded by nociceptive genes in the peritoneum of women with chronic pelvic pain (CPP) with or without endo-metriosisandtoinvestigatewhetherestrogensalternociceptivegeneexpressioninhumansensory neurons. Design: The study was performed using human tissue analysis and cell culture. Setting: The study was conducted at a university research institute. Patients: PeritonealbiopsieswereobtainedfromwomenwithCPPandendometriosis(n (cid:2) 12),CPP and no endometriosis (n (cid:2) 10), and no pain or endometriosis (n (cid:2) 5). Endometriosis lesions were obtained from women with endometriosis (n (cid:2) 18). Main Outcome Measures: mRNAs encoding ion channels ( P2RX3 , SCN9A , SCN11A , TRPA1 , TRPV1 ) and the neurotransmitter TAC1 were measured in human tissue samples and in human embryonic stem cell-derived sensory neurons treated with estrogens. Results: TRPV1, TRPA1 , and SCN11A mRNAs were significantly higher in the peritoneum from women with endometriosis ( P (cid:3) .001, P (cid:3) .01). TRPV1, SCN9A , and TAC1 were elevated in endometriosis lesions ( P (cid:3) .05). P2RX3 mRNA was increased in the peritoneum of women with CPP, with and without endometriosis ( P (cid:3) .05). Incubation of sensory neurons with 17 (cid:2) -estradiol increased TRPV1 mRNA ( P (cid:3) .01). The estrogen receptor- (cid:2) -selective agonist 2,3-bis(4-hydroxy-phenyl)-pro-pionitrile increased concentrations of TRPV1, P2RX3, SCN9A , and TAC1 mRNAs. Conclusions: Estrogen-dependentexpressionofTRPV1insensoryneuronsmayexplainwhyovarian suppression can reduce endometriosis-associated pain. Strategies directly targeting ion channels may offer an alternative option for the management of CPP. ( J Clin Endocrinol Metab 99: E1738–E1743, 2014)

associated pain. Women with endometriosis often suffer from cyclical pain, and pain may occur in association with other conditions, such as irritable bowel syndrome and migraine (2,3). Visceral, mechanical, and generalized hypersensitivity is extremely common in women with endometriosis (4), although there is no correlation between the extent of disease and reported pain scores (5). Stratton and Berkley (6) recently provided an overview of evidence for recruitment of nerves to endometriosis lesions, and they suggested that their stimulation and subsequent dialogue with the central nervous system may explain why women with endometriosis suffer from chronic pain.
Studies in patients with altered pain responses due to rare conditions as well as genetically modified mice (7) have identified key genes expressed by nerves that detect noxious stimuli (nociceptors). These include the TRP channel family (TRPA1, TRPV1), sodium voltage-gated ion channels (SCN9A, SCN11A) and purinergic X (P2X) receptors. An increase in the expression of neuropeptides, neurotrophins, and ion channels leading to an increased responsiveness of nociceptors has been suggested as an important contributor to development of hypersensitivity (8). For example, the TRPV1 channel (capsaicin receptor) has been implicated in development of hypersensitivity to heat associated with inflammation in chronic pain conditions (9). Recently studies have reported that TRPV1 immunoreactivity is increased in the peritoneum of women with CPP (10) and in lesions of women with endometriosis (11). The neuropeptide substance P (encoded by the TAC1 gene) has been immunolocalized to endometriosis lesions (12), and TRPA1 has been implicated in thermal and mechanical hypersensitivity to pain (13). The purinergic X family of receptors are voltage-gated ion channels that open in response to ATP released from cells during inflammation and have also been associated with mechanical hyperalgesia (14). Sodium channels are also critical in producing an acute response to noxious stimuli; Nav1.7 (SCN9A) and Nav1.9 (SCN11A) are both reported to be increased in a range of inflammatory pain conditions (7). Notably, induction of a hypoestrogenic state often ameliorates CPP, and although the mechanisms responsible for this remain poorly understood, estrogens are reported to influence the sensitivity of peripheral sensory neurons and central neuronal activity in pain states (2).
In this study we measured the concentrations of mR-NAs encoded by genes known to regulate nociception in tissue samples from women with CPP, some of whom were diagnosed with endometriosis. We also explored the impact of estrogens on their expression using a newly established in vitro model of human embryonic stem (ES) cellderived sensory neurons.

Patients and recovery of tissue samples
The study was approved by the Lothian Research Ethics Committee (LREC 11/AL/0376). Peritoneal biopsies [ϳ1.0 ϫ 0.5 cm size: stored in RNAlater (Applied Biosystems) at Ϫ20°C] were obtained from 3 groups of women (aged 18 -45 y) at the time of surgery with informed consent: 1) women undergoing laparoscopic investigation for CPP with histological evidence of pelvic endometriosis (n ϭ 12); 2) women undergoing laparoscopic investigation for CPP without evidence of endometriosis (n ϭ 10); and 3) women undergoing laparoscopic sterilization without CPP and without evidence of endometriosis (n ϭ 5). Endometriosis lesions were also collected from women with endometriosis (n ϭ 18). Peritoneal biopsies from women without endometriosis were recovered from a site prone to endometriosis (pouch of Douglas) and in women with endometriosis the peritoneal biopsy was recovered from a site adjacent to the endometriosis lesion. None of the women had taken exogenous hormones for at least 3 months at the time of sampling.

Calcium assay
Neurons were stimulated with 4 nM capsaicin (Sigma), and intracellular calcium was measured using a calcium indicator kit (BD Biosciences) with calcium flux captured using a NOVOstar microplate fluorometer (BMG Labtech).

Statistical analysis
Quantitative PCR (QPCR) data were analyzed using a oneway ANOVA and a Newman Keuls post hoc multiple comparison test. Statistics were generated using GraphPad Prism 6 software.

mRNAs encoding nociceptive proteins exhibit altered expression patterns in the peritoneum of women with CPP
Measurement of mRNAs encoded by TAC1 and nociceptive ion channels revealed differences between tissue samples; TAC1 was elevated only in endometriosis lesions ( Figure 1A; P Ͻ .05), and P2RX3 was increased in the peritoneum of women with CPP, regardless of whether they were diagnosed with endometriosis compared with the peritoneum of healthy women (Figure 1B; P Ͻ .05). SCN9A (Nav1.7) was elevated in lesions from women with endometriosis ( Figure 1C; P Ͻ .05), whereas SCN11A was significantly higher in the peritoneum of women with CPP and endometriosis compared with the peritoneum of women with CPP alone ( Figure 1D; P Ͻ .001). mRNA encoded by TRPA1 was significantly increased in the peritoneum of women with endometriosis compared with the peritoneum of healthy women and those with CPP alone ( Figure 1E; P Ͻ .001). TRPV1 was elevated in peritoneum (P Ͻ .01) and in lesions (P Ͻ .05) of women with endometriosis compared with the peritoneum of healthy women ( Figure 1F).

mRNA concentrations of nociceptive genes are regulated by estrogens in ES cell-derived sensory neurons
Sensory neurons immunopositive for neurofilament were generated in vitro (15) (Figure 2, A-C). Differentiation was accompanied by a reduction in octamer transcription factor-4 mRNA concentrations and an increase in TAC1, SCN9A, and SCN11A ( Figure 2D), consistent with differentiation of sensory neurons with a nociceptorlike phenotype. When the differentiated cells were incubated with capsaicin (activates TRPV1), an intracellular calcium flux was detected ( Figure 2E). The expression of ER subtypes changed during differentiation; ER␣ mRNA concentrations decreased and ER␤ mRNAs increased ( Figure 2F). Stimulation of sensory neurons with the estrogen ligands E2, DPN, and PPT revealed ER-dependent regulation of nociceptive ion channels. TAC1 and P2RX3 mRNAs were increased after the incubation with DPN compared with cells incubated with DPN in the presence of ICI ( Figure 2G; P Ͻ .001) or vehicle control (DMSO;   Figure 2H; P Ͻ .05), respectively. TAC1 was also elevated by PPT compared with PPT ϩ ICI ( Figure 2G; P Ͻ .01). Although SCN9A increased with DPN (P Ͻ .05), this was not blocked by the addition of ICI (Figure 2I), and the expression of SCN11A ( Figure 2J) and TRPA1 ( Figure  2K) did not appear to be ER dependent. In contrast, TRPV1 mRNA concentrations were elevated by both E2 (P Ͻ .05) and DPN ( Figure 2I; P Ͻ .01), and this effect was abrogated by ICI.

Discussion
We believe this is the first study to detect differences in the expression of mRNAs encoded by the ion channels TRPA1, P2RX3, SCN9A, and SCN11A in the peritoneum and lesions of women suffering from CPP, some of whom had endometriosis. We believe these insights may be indicative of sensitization of nerves present within the peritoneal lining of women with pain and therefore shed new light on the mechanisms responsible for development of chronic hypersensitivity.
In women with active endometriosis, we discovered that the concentrations of mRNAs encoded by SCN11A, TRPA1, and TRPV1 were all significantly increased in the samples of peritoneum. We believe these novel findings are consistent with nociceptive changes taking place within the peritoneum of women with endometriosis, suggesting an increased sensitization of the peritoneum adjacent to the lesion. We detected significant increases in TRPV1 mRNAs in both the peritoneum and in lesions, confirming and extending a previous study (11). Notably, P2RX3 mRNAs were significantly elevated in the peritoneum of women with CPP, regardless of whether they had active endometriosis and would be consistent with a neuropathic component to idiopathic CPP.
To date, researchers have used rat or chick dorsal root ganglia (DRG) neurons as an in vitro model to investigate the function of sensory neurons. However, because DRG neurons represent a mixed population of somatosensory neurons, the development of a differentiation protocol allowing the generation of a population of human sensory neurons with a nociceptor-like phenotype (15) is particularly welcome. In the current study, we detected expression of estrogen receptor mRNAs in our ES cell-derived cultures. Treatment of cells with E2 or the ER␤ agonist DPN induced a significant increase in TRPV1 mRNAs that was blocked by the addition of the ER antagonist ICI.
In addition, P2RX3, SCN9A, and TAC1 mRNAs were all increased when cells were incubated with the ER␤ agonist DPN, and TAC1 was also increased by PPT. We have previously demonstrated DPN-dependent changes in gene expression in endometrial endothelial cells and have shown that these were mediated via specificity protein-1 tethered transcription (16). Bioinformatics revealed the presence of specificity protein-1 transcription factor binding sites in the promoter regions of TRPV1, P2RX3, SCN9A, and TAC1 (Match; Biobase international.com). TRPV1 mRNA was the only transcript that was increased in response to E2 treatment and abrogated by the antiestrogen ICI. Notably, bioinformatics detected several estrogen response elements in the promoter region of TRPV1 and throughout the length of the transcript (Dragon estrogen response element locator version 6.0; National Institute of Health) that could facilitate direct binding of E2-activated ER receptor dimers. Further studies are required to validate these findings.
Estrogens are reported to have both pro-and antinociceptive properties. Mice exhibit hyperalgesia when estrogens are depleted after ovariectomy (17), and nociceptive responses are lower in ER␤ knockout mice during the early stages of inflammation (18). The results of our study suggest estrogens may have a direct impact on nociception by up-regulating the expression of ion channels and would be consistent with reports that the expression of Trpv1 and P2x3 proteins are decreased in DRGs from mice with targeted deletions of Esr2 and Esr2 (19).
In summary, the results of this study indicate that therapies targeting P2RX3 may be useful in treating CPP in women with a diverse range of painful etiologies, whereas women with active endometriosis could benefit from treatments targeting TRPV1, TRPA1, SCN9A, or SCN11A. Studies using sensory neurons suggest regulation of TRPV1 by estrogens may provide an explanation for the reduction in pain experienced when ovarian steroids are suppressed and an opportunity for development of novel therapies.