A link between plasma membrane calcium ATPase 2 (PMCA2), estrogen and estrogen receptor α signaling in mechanical pain

Earlier studies on genetically modified mice indicated that plasma membrane calcium ATPase 2 (PMCA2), a calcium extrusion pump, plays a novel and sex-dependent role in mechanical pain responses: female, but not male, PMCA2+/− mice manifest increased mechanical pain compared to female PMCA2+/+ mice. The goal of the present studies was to determine the contribution of ovarian steroids to the genotype- and sex-dependent manifestation of mechanical pain in PMCA2+/+ versus PMCA2+/− mice. Ovariectomy increased mechanical pain sensitivity and 17β-estradiol (E2) replacement restored it to basal levels in PMCA2+/+ mice, but not in PMCA2+/− littermates. Intrathecal administration of an estrogen receptor alpha (ERα) agonist induced ERα signaling in the dorsal horn (DH) of female PMCA2+/+ mice, but was ineffective in PMCA2+/− mice. In male PMCA2+/+ and PMCA2+/− mice, E2 treatment following orchidectomy did not recapitulate the genotype-dependent differential pain responses observed in females and the agonist did not elicit ERα signaling. These findings establish a novel, female-specific link between PMCA2, ERα and mechanical pain. It is postulated that PMCA2 is essential for adequate ERα signaling in the female DH and that impaired ERα signaling in the female PMCA2+/− mice hinders the analgesic effects of E2 leading to increased sensitivity to mechanical stimuli.


Results
Effects of ovariectomy (OVX) and E2 replacement on mechanical and heat sensitivity in PMCA2 +/+ and PMCA2 +/− mice. To determine whether circulating ovarian hormones modulate mechanical pain in a genotype-dependent manner and to assess whether E2 reverses the possible effects of OVX, we performed the von Frey filament test on SHAM-operated and ovariectomized PMCA2 +/+ and PMCA2 +/− mice treated with vehicle or E2. Mechanical sensitivity was evaluated at 2 weeks postoperatively. The paw withdrawal threshold was significantly decreased in ovariectomized PMCA2 +/+ mice treated with vehicle as compared to the SHAM-operated PMCA2 +/+ littermates (p = 0.021). Paw withdrawal thresholds in E2-treated ovariectomized PMCA2 +/+ mice were significantly higher than vehicle-treated ovariectomized PMCA2 +/+ mice (p = 0.039) and not different from SHAM values (Fig. 1A). In contrast, paw withdrawal thresholds in vehicle-or E2-treated ovariectomized PMCA2 +/− mice were not significantly different than those observed in SHAM-operated PMCA2 +/− controls (Fig. 1A). These studies indicated that depletion of circulating ovarian hormones affects pain sensitivity in PMCA2 +/+ mice, but does not alter pain responsiveness in PMCA2 +/− mice. Moreover, in ovariectomized PMCA2 +/+ mice, E2 replacement was sufficient to restore paw withdrawal thresholds to SHAM values whereas it was ineffective in ovariectomized PMCA2 +/− mice. Of note, in agreement with our earlier report showing increased mechanical sensitivity in the intact female PMCA2 +/− compared to the intact PMCA2 +/+ mice 4 , SHAM-operated PMCA2 +/− mice were more sensitive to a mechanical stimulus than SHAM-operated PMCA2 +/+ mice (p = 0.0005).
To determine whether the effects of OVX and E2 replacement in PMCA2 +/+ and PMCA2 +/− mice are modality-specific, we assessed heat sensitivity in the same mice using the Hargreaves' plantar thermal test. There were no statistical differences in paw withdrawal latencies among SHAM, vehicle-treated ovariectomized, and E2-treated ovariectomized PMCA2 +/+ or PMCA2 +/− mice (Fig. 1B). These findings indicate that in our experimental paradigm, the effects of circulating ovarian steroids on pain sensitivity are modality-dependent.
Expression of ERα in the DH of female PMCA2 +/+ and PMCA2 +/− mice. Because OVX and E2 replacement affected mechanical pain sensitivity in PMCA2 +/+ mice but were ineffective in PMCA2 +/− littermates, we investigated whether this is due to decreased estrogen receptor expression in the DH of PMCA2 +/− mice. We focused on ERα because of its well-established role in in pain responses and mechanisms in the DH 28,35,36 . We examined ERα protein expression in the DH of the PMCA2 +/+ and PMCA2 +/− mice. Three immunoreactive bands at molecular weights 66, 52, and 32 KDa were detected when the western blots were probed with an antibody against ERα ( Fig. 2A). Quantification of the band intensities indicated no significant differences between the samples obtained from the PMCA2 +/+ and PMCA2 +/− mice (Fig. 2B). In addition, we did not find any significant difference in the ERα transcript levels in the DH of the PMCA2 +/+ and PMCA2 +/− mice (Fig. 2C). Note that the 32 KDa band was the most abundant and was detected only after 5 min exposure whereas the 66 and 52 KDa bands were detected when the blots were exposed for 60 min.
ERα signaling is impaired in the DH of female PMCA2 +/− mice. We next investigated a potential impairment of ERα signaling in the DH of female PMCA2 +/− mice. Activation of ERα signaling can lead to activation of ERK/MAPK, JNK/MAPK, or p38/MAPK signal transduction pathways, resulting in the phosphorylation of ERK, JNK, and p38. We first assessed whether basal phospho-ERK1/2 (pERK1/2) and total ERK, phospho-JNK (pJNK) and JNK, phospho-p38 (pp38) and total p38 levels in the naïve, intact PMCA2 +/+ and PMCA2 +/− are comparable (Fig. 3A,B). Basal pERK and total ERK as well as pJNK and total JNK levels were comparable in PMCA2 +/+ and PMCA2 +/− mice. pp38 levels were low and could not be reliably analyzed (results not showed). We then assessed whether intrathecal administration of PPT, an ERα agonist, increases pERK/ERK and pJNK/JNK ratios. Administration of PPT significantly increased pERK levels in PMCA2 +/+ mice as compared to vehicle treated PMCA2 +/+ mice ( Fig. 3C; p < 0.01). In contrast, intrathecal PPT did not increase pERK levels in the PMCA2 +/− mice compared to vehicle-treated PMCA2 +/− mice raising the possibility of impairment in ERα signaling (Fig. 3C). pJNK/JNK ratios were not significantly different in PPT-or vehicle-treated mice of both genotypes (Fig. 3D). Even after PPT treatment, pp38 levels were not adequately detected, were variable and could not be analyzed reliably (results not shown). Depletion of testicular hormones combined with E2 treatment does not recapitulate the female pattern of genotype-dependent mechanical sensitivity in male PMCA2 +/− and PMCA2 +/+ mice. We previously reported that gonadally intact female PMCA2 +/− mice exhibit higher  mechanical sensitivity compared to female PMCA2 +/+ mice whereas there are no differences in mechanical pain responses in the gonadally intact male PMCA2 +/− and PMCA2 +/+ mice 4 . We undertook investigations to determine whether orchidectomized male PMCA2 +/+ and PMCA2 +/− mice treated with E2 exhibit the female pattern of genotype-dependent differential mechanical pain responses. PMCA2 +/+ and PMCA2 +/− mice underwent orchidectomy (ORX) followed by implantation of silastic capsules containing E2 or vehicle as described in methods. Two weeks after ORX and the implantation of the capsules, the von Frey filament test and Hargreaves' plantar thermal test were performed. Orchiectomy significantly decreased the paw withdrawal thresholds in both PMCA2 +/+ (p < 0.0001) and PMCA2 +/− (p < 0.0001) mice, indicating that depletion of testicular hormones increases mechanical sensitivity independent of genotype and suggesting that testicular hormones in the gonadally intact mice have analgesic effects. There were no significant differences in paw withdrawal thresholds between vehicle-treated or E2-treated orchidectomized mice in either genotype (Fig. 4A). Thus, E2 treatment of gonadectomized male PMCA2 +/+ and PMCA2 +/− mice does not recapitulate the mechanical pain sensitivity pattern observed in female PMCA2 +/+ and PMCA2 +/− mice. When heat sensitivity was assessed by the Hargreaves' test, there were no statistical differences in paw withdrawal latencies between SHAM, vehicle-treated orchidectomized, and E2-treated orchidectomized PMCA2 +/+ or PMCA2 +/− mice demonstrating that the effects of circulating testicular hormones on pain sensitivity are modality-dependent (Fig. 4B).
PPT does not activate ERα signaling in male PMCA2 +/+ and PMCA2 +/− mice. Because E2 treatment did not elicit changes in paw withdrawal threshold of orchidectomized PMCA2 +/+ or PMCA2 +/− mice, we investigated whether ERα signaling in the DH of male PMCA2 +/+ or PMCA2 +/− mice is different than in female PMCA2 +/+ and PMCA2 +/− mice. After confirming equivalent ERα protein and mRNA expression in the DH of gonadally intact male PMCA2 +/+ and PMCA2 +/− mice ( Fig. 5A-C), we treated the mice with PPT utilizing the same experimental conditions as the females. In contrast to female mice, intrathecal PPT did not induce phosphorylation of ERK. We did not find significant differences in pERK/ERK or pJNK/JNK ratios in the DH of

Discussion
The present study shows, for the first time, that mechanical pain responses in the female PMCA2 +/+ and PMCA2 +/− mice are modulated by circulating ovarian steroids in distinct manner. The increase in mechanical pain sensitivity in PMCA2 +/+ mice following OVX suggests that ovarian sex hormones play an analgesic role in the gonadally intact PMCA2 +/+ mice. Such analgesic effects have been documented in a number of experimental paradigms 22,23,25,43 . Our studies also demonstrate that E2 replacement alone is sufficient to counteract the effects of OVX in PMCA2 +/+ mice. In contrast, OVX and E2 replacement do not alter mechanical pain responses in PMCA2 +/− littermates. This is likely due, at least partly, to the impairment of ERα signaling in the DH of the female PMCA2 +/− mice as indicated by the absence of ERK activation in response to an ERα agonist. Our investigations focused on ERα even though expression of ERβ and GPR30 in the DH has been reported 41,42,44,45 . ERα was of interest because its role in pain responses is well documented 28,35,36 , whereas studies on ERβ and GPR30, in relation to pain, are limited. Our studies unraveled the presence of several ERα immunoreactive bands in the DH. The 66 KDa band corresponds to the molecular weight of classical ERα, whereas the 52 KDa immunoreactive band, possibly an ERα isoform, has also been observed in the hippocampus 46 . The most abundant band at 32 KDa could be a variant of the recently described membrane-bound ERα-36, which also signals through activation of ERK 47,48 . Similar to what was observed in the hippocampus 46 , the abundance of the 66 KDa band was lower than the other two immunoreactive bands, raising the possibility that the 32 and 52 KDa variants might play a more prominent role than the classical ERα in the DH.
How can reduced PMCA2 expression in female PMCA2 +/− mice alter pain processing in the DH such that the effects of E2 on pain sensitivity are different than that observed in the wild type? We postulated that low PMCA2 expression in DH neurons of female PMCA2 +/− mice and the consequent calcium dyshomeostasis alter molecular mechanisms underlying mechanical pain leading to reduced responsiveness of neurons to E2. In agreement with this postulate, we found that ERα signaling is impaired in the DH of the female PMCA2 +/− mice. It is important to note that a role for calcium in the activation of ERα signaling has been documented 49 . In the absence of adequate ERα signaling, the analgesic effects of E2 might be compromised, resulting in higher mechanical pain sensitivity in the female PMCA2 +/− mice. Since genotype-dependent differences in mechanical pain are only observed in female PMCA2 +/+ and PMCA2 +/− mice, we assessed whether feminization of male PMCA2 +/+ and PMCA2 +/− mice by ORX combined with E2 treatment recapitulates the mechanical pain differences observed female PMCA2 +/+ and PMCA2 +/− mice. ORX alone, increased mechanical sensitivity in both PMCA2 +/+ and PMCA2 +/− mice indicating that testicular hormones exert analgesic effects irrespective of genotype. Treatment of orchidectomized PMCA2 +/+ and PMCA2 +/− mice with E2 did not recapitulate the differential mechanical pain sensitivity observed in female PMCA2 +/+ and PMCA2 +/− mice. It is possible that the mechanisms underlying mechanical pain responses in the male and female differ, with females being more susceptible to changes in PMCA2 expression and males being more refractive to alterations in PMCA2 expression and function. Interestingly, a study analyzing sex differences in signaling at hippocampal inhibitory synapses, reported that despite the presence of ERα-mGluR1 complexes in the hippocampus of both female and male rats, the regulation of this complex by E2 occurred only in females. Consequently, the E2-elicited changes in downstream signaling were enhanced in females as compared to males 50 . It is also conceivable that activation of ERα signaling in the male DH necessitates an E2 treatment regimen different than what was employed in our investigations, which was particularly devised for female mice. Finally, E2 effects in the DH of male and female mice could be mediated by different estrogen receptors.
Our studies revealed effects of ovarian hormones on pain sensitivity in a modality-specific manner, since OVX increased mechanical but not heat sensitivity in the wild type mouse. Other investigators have also shown that heat sensitivity is not altered at 2 weeks post-OVX in the C57Bl/6 mouse 25 . However, in the study performed by Sanoja and Cervero 25 , increased heat sensitivity manifested at 4 weeks post-OVX, whereas we did not observe any changes in heat sensitivity even at 4 weeks post-OVX (unpublished results). However, it is important to note that Sanoja and Cervero 25 used the hot plate paw withdrawal test whereas we utilized the Hargreaves' plantar thermal test. These two tests differ because the Hargreaves' test is considered to reflect primarily spinally mediated pain processing whereas the hot plate paw withdrawal test engages both spinal and supraspinal mechanisms 51,52 . In an additional report, increased heat sensitivity was observed in rats at 2 weeks post-OVX utilizing the tail flick thermal test 53 . The discrepancy between our results and this latter study could be due to the use of distinct animal species and behavioral tests. It is important to mention that many animal studies investigated sex specificity in pain responses. The outcomes of these investigations vary and are often contradictory. This may be a result of using different animal strains, experimental approaches, and experimental sample size 54,55 . The mice utilized in our experiments were on a C57Bl/6 background. However, when we assessed heat and mechanical sensitivity in PMCA2 +/− and PMCA2 +/+ mice on a Black Swiss background we obtained similar results (unpublished data). Therefore, the sex bias in genotype-dependent mechanical pain responses is not strain dependent.
Based on our findings, we propose that there is relationship between PMCA2 expression and ERα signaling in the DH of female mice and that a reduction in PMCA2 levels in female PMCA2 +/− mice interferes with ERα signaling and abrogates the analgesic effects of E2 which are observed in the female PMCA2 +/+ mice. This leads to increased mechanical pain sensitivity in the female PMCA2 +/− mice. In contrast, male mice might utilize different mechanisms which are similarly modulated by testosterone in both genotypes and are not dependent on E2.

Methods
Animals. C57Bl/6 strain male and female PMCA2 +/+ and PMCA2 +/− mice 56 were bred in house. All mice were housed in a barrier facility at Rutgers-New Jersey Medical School on a 12:12 h light-dark cycle at an ambient temperature of 22-23 °C with water and standard chow provided ad libitum. All procedures were approved by the Rutgers University Institutional Animal Care and Use Committee, were carried out in accordance with the relevant guidelines and regulations, and were consistent with the ethical guidelines of the US National Institutes of Health and the International Association for the Study of Pain. All efforts were made to minimize animal suffering and to limit the number of animals tested.
OVX and E2 Replacement. Two-month-old female PMCA2 +/+ and PMCA2 +/− mice were anesthetized with a mixture of ketamine (80 mg/kg; Vedco, St. Joseph, MO, USA) and xylazine (10 mg/kg; Akorn Inc., Decatur, IL, USA), administered intraperitoneally (i.p.). In ovariectomized mice, both ovaries were surgically removed whereas the SHAM mice underwent the retroperitoneal surgery without removal of ovaries. A 3 cm dorsal, midline incision was made in order to separate the skin from the underlying fascia. Another incision was made through the fascia and the adipose tissue that surrounds the ovary was isolated. Once the ovaries were identified, the uterine horns were ligated, the ovaries were removed and the surgical bed was cauterized to prevent bleeding. The uterine horn and the adipose tissue were returned to the retroperitoneal cavity and the wound was sutured. The protocol for E2 replacement was adapted from Ström et al. 57 . Two cm long silastic capsules (Dow Corning; 0.063in I.D; 0.125in O.D.; Fischer Scientific; Pittsburgh, PA, USA) were filled either with E2 (Sigma; E8875; St. Louis, MO, USA) dissolved in sesame oil (Sigma) at a concentration of 25 µg/ml or sesame oil (vehicle) (approx. 0.05 mL). The ends of the capsules were plugged with 3 mm long wooden sticks. The capsules were incubated overnight in their respective solutions, either E2 in sesame oil or sesame oil. Subsequently, they were subcutaneously implanted at the nape of the neck immediately following OVX or SHAM surgery. For postoperative care, the mice received a single subcutaneous injection of Buprenorphine (0.05 mg/kg; Hospira, Lake Forest, IL, USA) on the day of the surgery. To ascertain successful OVX, uterine weights were determined 2 weeks postoperatively, at the time when mice were euthanized. Atrophy of the uterus following the depletion of circulating ovarian hormones has previously been shown to be a reliable index of successful OVX 53 . An increase in uterine weight was used as a confirmation of successful E2 replacement. The uterine weight (mg) for the PMCA2 +/+ mice groups were as follows: SHAM-operated (67.7 ± 13.6, n = 10), ovariectomized treated with vehicle (15.8 ± 1.79, n = 10), and ovariectomized treated with E2 (116.7 ± 18.8, n = 11). Likewise, the uterine weight (mg) for the PMCA2 +/− mice groups were as follows: SHAM-operated (39.3 ± 2.12, n = 12), ovariectomized treated with vehicle (13.5 ± 1.04, n = 12), and ovariectomized treated with E2 (75.4 ± 8.82, n = 11).
ORX and E2 treatment. Two-month-old male PMCA2 +/+ and PMCA2 +/− mice were anesthetized. In orchidectomized mice both testes were removed whereas SHAM mice underwent the surgery without removal of the testis. A transverse incision was made at the abdomen and the skin was separated from the muscle. Another transverse incision was made through the muscle and the testicular fat pad was pulled through the incision. The testes were removed and the surgical bed was cauterized. 17β-estradiol replacement and postoperative care were performed as described above. Successful ORX was confirmed by measuring the weight of the seminal vesicles at 2 weeks post-ORX. Atrophy of seminal vesicles, as a result of circulating testicular hormone depletion, has been shown to be a reliable index of successful ORX 58   PMCA2 +/− mice groups were as follows: SHAM-operated (172.7 ± 7.14, n = 19), orchidectomized treated with vehicle (15.5 ± 1.10, n = 17), and orchidectomized treated with E2 (18.6 ± 2.28, n = 19).
Behavioral Analyses for the assessment of pain modalities. Thermal and/or mechanical sensitivity were assessed 2 weeks following OVX and ORX. The Plantar Thermal Test (Hargreaves' Method; intensity of 15) and the von Frey filament test (up/down paradigm) were performed as previously described 4 . The evaluator was blinded to experimental conditions.

Intrathecal delivery of the ERα agonist PPT by lumbar puncture (LP). 4,4′,4′′-(4-Propyl-[1 H]-pyrazole-
1,3,5-triyl) trisphenol; PPT; Tocris; Minneapolis, MN, USA) was dissolved in dimethyl sulfoxide (DMSO), and conserved in aliquots at −80 °C, until use. To perform LPs, mice were anesthetized with isoflurane (1.0 l/min at a concentration of 3.0% in oxygen). A 27-gauge needle attached to a Hamilton syringe was inserted percutaneously into the subarachnoid space via the interlaminar space between L5 and L6. Injections into the intrathecal space were performed caudal to the conus medullaris to avoid damage to the spinal cord. Ten ng PPT was administered in a total volume of 6 μl DMSO. Control mice received 6 μl of DMSO (vehicle). The needle was maintained in place for 30 s before it was slowly removed to prevent egress of the infiltrate from the puncture site. Mice were sacrificed 1 h following administration of PPT or vehicle, and the lumbar DH was immediately removed and frozen on dry ice. . Five to 20 µg of total protein were loaded in each lane of a 4-12% Bis-Tris gel or 4-15% stain-free Tris-Glycine eXtended (TGX) gel (Bio-Rad). Electrophoresis was performed for 1 h at 200 V. The protein was then electrotransferred onto a polyvinylidene difluoride (PVDF) membrane for 1 h at 100 V for high molecular weight proteins or for 15 min at 25 V for small molecular weight proteins using the Trans-Blot Turbo transfer system (Bio-Rad). The membranes were stained with Ponceau S Staining Solution (Sigma) according to the manufacturer's instructions and blocked for 60 min with buffer consisting of 5% bovine serum albumin (BSA) in Tris-buffered saline (TBS; 20 mM Trizma Base and 500 mM NaCl, pH 7.5) containing 0.1% Tween-20 (T-TBS). Membranes were probed overnight at 4°C, with primary antibodies against phospho-p44/42 MAPK (ERK1/2; Thr202/Tyr204; Cell Signaling, Danvers, MA, USA; 1: 20,000), phospho-SAPK/JNK (Thr183/Tyr185; Cell Signaling; 1:5000), and phospho-p38 MAPK (Thr180/Tyr182; Cell Signaling; 1:1000) diluted in 5% BSA containing 0.1% T-TBS. The phosphorylated proteins were stripped off by incubating membranes in a buffer containing 1 M Tris-hydrochloride (pH 6.8), 20% sodium dodecyl sulfate (SDS) and β-mercaptoethanol in deionized water at 50 °C for 30 min. The membranes were then re-probed with antibodies against total p44/42 MAPK (ERK1/2; Cell Signaling; 1: 20,000), total SAPK/JNK (Cell Signaling; 1:5000), and total p38 MAPK (Cell Signaling; 1:5000) overnight at 4 °C.
For the evaluation of ERα, membranes were blocked for 60 min with a buffer consisting of 5% non-fat milk dissolved in 0.1% T-TBS and then probed overnight at 4 °C with primary antibodies against ERα (Abcam; Cambridge, MA, USA; 1:2000) diluted in 5% BSA containing 0.1% T-TBS. Bands were visualized using Clarity (Bio-Rad) and the Chemidoc system (Bio-Rad). Signal intensity was quantified using the UN-SCAN-IT software (Silk Scientific, Orem, UT, USA) and normalized to total protein detected in the stain-free gel utilizing the Chemidoc system. Quantitative real time-PCR. RNA extraction and quantitative real time-PCR (qRT-PCR) were performed as previously described 4 using 100 ng cDNA and 100 nM primers and SYBR Green master mix with Rox (Clontech, Mountain View, CA, USA). Samples were amplified for 40 cycles using an Applied Biosystems Realplex. The amplification conditions were as follows: denaturation 95°C, 10 sec; annealing: 58°C, 45 sec; extension: 72 °C, 30 sec. The samples were then heated at 95°C for 15 sec and subsequently underwent a melting curve analysis from 60°C to 95°C while measuring fluorescence at every degree increase in temperature. Quantification was performed using the cycle threshold method 59 with GAPDH as an internal control and results were presented as percent of their respective control groups. Primer sequences (5′ to 3′; Real Time Primers, Elkins Park, PA, USA) are provided below for ERα (ESR1) and GAPDH. independent t-test with a Bonferroni correction was used to determine differences between different factors and individual groups. Data are presented as the mean ± standard error of the mean (S.E.M).

Data Availability
The data sets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.