Estrogens influence female itch sensitivity via the spinal gastrin-releasing peptide receptor neurons

Significance Many women exhibit a dramatic increase in itch during pregnancy, but the underlying mechanism is unknown. Here, we demonstrate that the female sex steroid hormone estradiol, but not progesterone, enhances itch-related scratching behavior in female rats elicited by histamine, the prototypical itch mediator in humans. This is associated with an enhancement in histamine-evoked activity of a subset of spinal dorsal horn neurons that express a neuropeptide receptor, gastrin-releasing peptide receptor (GRPR), that was previously shown to be involved in spinal cord processing of itch. These findings may account for why itch sensation varies with estrogen levels and provide a basis for treating histamine-related itch diseases in females by targeting GRPR.

I tch and pain warn the body of potential tissue damage and are indispensable to survival. The threshold and intensity of itch and pain are not absolute but depend on various environmental and psychological factors such as stress, mood, and anxiety (1)(2)(3). Female-specific pruritus occurs during pregnancy and menopause when circulating female hormones dramatically fluctuate (4,5). Approximately 20% of pregnant women have itch symptoms during pregnancy (6), such as pruritic urticarial papules and plaques of pregnancy, intrahepatic cholestasis of pregnancy, and pemphigoid gestationis. Female patients with atopic dermatitis often experience atopic eruption and cutaneous deterioration associated with their pregnancy (7)(8)(9). Although these itch symptoms impair female quality of life, the mechanism is still unknown. Histamine is an important peripheral itch mediator. Once released from mast cells activated by irritant stimuli and allergens, histamine induces not only inflammation but also itch triggered by the excitation of a subset of unmyelinated C fibers (10). Centrally, neurons expressing the gastrin-releasing peptide (GRP) and its receptor (GRPR) play a critical role in the spinal cord transmission of itch signals independent of pain (11)(12)(13)(14)(15). In the present study, we focused on the female sex hormones as candidates for the cause of itch in women. We discovered that female itch sensitivity is induced by estrogens via the spinal GRP/ GRPR system in rats.

Results
To determine the effects of female sex hormones on itch sensitivity, we first investigated histamine-evoked hind paw scratching as a marker of itch behavior (Fig. 1A) in male and female Wistar rats as well as in bilaterally ovariectomized (OVX) females implanted with a blank (b) capsule as a control (OVX + b), OVX females treated with a physiological level of 17β-estradiol (E) (OVX + E), OVX females treated with a physiological level of progesterone (P) (OVX + P), and OVX females treated with both estradiol and progesterone (OVX + EP) (SI Appendix).
Estradiol, but not progesterone, replacement strikingly induced a significant elevation in the number of scratch bouts ( Fig. 1 B and E and SI Appendix, Fig. S1: Kruskal-Wallis post hoc test: *P < 0.0083, b versus E; **P < 0.001, E versus P and E versus EP) and prolonged the duration ( Fig. 1 C and F: Kruskal-Wallis post hoc test: *P < 0.0083; b versus E, b versus EP, E versus P, **P < 0.001; E versus EP) of histamine-evoked scratching behavior. Estradiol also shortened the latency of histamine-evoked scratching ( Fig. 1D: Kruskal-Wallis post hoc test: *P < 0.0083, E versus EP). A significant correlation was observed between the number of scratch bouts and the total duration of scratching/ 60 min in all groups ( Fig. 1G: Pearson's correlation coefficient test: Male, r = 0.906, P = 0.000; Female, r = 0.889, P = 0.001; OVX + b, r = 0.960, P = 0.000: OVX + E, r = 0.990, P = 0.000; OVX + P, r = 0.975, P = 0.000; OVX + EP, r = 0.941, P = 0.000). We additionally observed a sex difference as well as estrogen enhancement of scratching elicited by another itch mediator, chloroquine (CQ) (SI Appendix, Fig. S2 A-H). Intradermal Significance Many women exhibit a dramatic increase in itch during pregnancy, but the underlying mechanism is unknown. Here, we demonstrate that the female sex steroid hormone estradiol, but not progesterone, enhances itch-related scratching behavior in female rats elicited by histamine, the prototypical itch mediator in humans. This is associated with an enhancement in histamine-evoked activity of a subset of spinal dorsal horn neurons that express a neuropeptide receptor, gastrin-releasing peptide receptor (GRPR), that was previously shown to be involved in spinal cord processing of itch. These findings may account for why itch sensation varies with estrogen levels and provide a basis for treating histamine-related itch diseases in females by targeting GRPR. injection of saline vehicle did not elicit significant scratching in any OVX group (SI Appendix, Fig. S3 A-C). In contrast, estradiol reduced sensitivity to innocuous von Frey mechanical stimuli ( Fig. 1H: Kruskal-Wallis post hoc test: *P < 0.0083, b versus EP, **P < 0.001; b versus E, E versus P, P versus EP). In addition, estradiol did not affect the thermal pain sensitivity (Fig. 1I). In summary, estradiol enhanced histamine-and CQ-evoked scratching behavior in OVX females while reducing mechanosensitivity. We next addressed whether estrogens affect the expression of itch and pain mediators in the spinal somatosensory system. GRP-and GRPR-expressing neurons are critical for itch signaling (11,12). In the spinal dorsal horn, estradiol replacement resulted in a significant increase in GRP messenger RNA (mRNA) compared with OVX + b group ( Fig. 2A: one-way ANOVA post hoc test: *P < 0.05, b versus E; **P < 0.01; b versus EP, P versus EP) as well as an increase in GRP protein expression ( Fig. 2B: one-way ANOVA post hoc test: **P < 0.01; b versus EP, P versus EP; Welch's t test: *P < 0.05, E versus P). The dorsal horn expression of GRPR mRNA (Fig. 2C) or GRPR protein (Fig. 2D) was not affected by estradiol replacement. There was no significant effect of estradiol or progesterone replacement on the expression of the mRNA for the histamine H1 receptor (Fig. 2E), the neurokinin-1 receptor (NK1R) (Fig. 2F) involved in spinal nociceptive transmission (16), or for Bhlhb5 expressed in spinal BI-5 itch-inhibitory interneurons ( Fig. 2G) (17).
We then investigated whether estrogen-dependent histaminergic itch involves the functional activation of GRPR-expressing neurons in the spinal cord. These GRPR neurons were localized in the spinal dorsal horn and caudal part of the spinal trigeminal nucleus, crucial relay points for itch transmission (SI Appendix, Fig. S4). In GRPR-mRFP1 transgenic rats, GRPR expression was neither observed in glial cells (i.e., astrocytes and microglia) ( Student's t test: † † P < 0.01, C versus I side of OVX + E; Student's t test: **P < 0.01, OVX + b versus OVX + E in the I side), and percentages ( Fig. 3O: Welch's t test: † † P < 0.01, C versus I side of OVX + b; Welch's t test: † † P < 0.01, C versus I side of OVX + E; Welch's t test: *P < 0.05, OVX + b versus OVX + E in the I side) in ipsilateral superficial laminae of the dorsal horn compared with the OVX + b group. Estradiol replacement by itself resulted in a marked increase in the number of GRPR-positive neurons ( Fig. 3L: Student's t test: **P < 0.01, OVX + b versus OVX + E in the C side; Student's t test: **P < 0.01, OVX + b versus OVX + E in the I side). In addition, intradermal injection of histamine in the hind paw of the OVX + E group resulted in a significant increase in hind paw biting behavior, which is thought to reflect itch (18) (Fig. 3 P and Q: biting number: Student's t test: P = 0.063; biting duration: Welch's t test: *P < 0.05, OVX + b versus OVX + E) with no change in latency to hind paw biting (Fig. 3R). Furthermore, intrathecal administration of the GRPR antagonist RC3095 dosedependently reduced histamine-evoked scratching behavior in the estrogen-treated females ( Fig. 3 S-V: scratching number: Mann-Whitney U test: *P < 0.05, control [aCSF + saline] versus histamine with aCSF, histamine with aCSF versus histamine with 0.1 nmol RC3095, histamine with aCSF versus histamine with 1 nmol RC3095; scratching duration: Welch's t test: P = 0.058, control versus histamine with aCSF; Mann-Whitney U test: *P < 0.05, histamine with aCSF versus histamine with 1 nmol RC3095; latency: Student's t test: *P < 0.05, control versus histamine with aCSF, histamine with aCSF versus histamine with 1 nmol RC3095). Using the "cheek" model (19) (Fig. 3W), histamine elicited hind paw scratch bouts indicative of itch in the estrogentreated group but very few ipsilateral forelimb wipes indicative of pain or grooming behavior ( Fig. 3X: Kruskal-Wallis post hoc test: † P < 0.016, scratching versus wiping, scratching versus grooming). Such scratching was significantly attenuated by the histamine H1 receptor antagonist terfenadine ( Fig. 3 X and Y: scratch bouts: Mann-Whitney U test: *P < 0.05; scratch duration: Student's t test: *P < 0.05, histamine versus histamine with terfenadine), and latency to scratching tended to increase following terfenadine ( Fig. 3Z: Welch's t test: P = 0.089). These data indicate that histamine-evoked itch is mediated by peripheral histamine H1 receptor and spinal GRPR-expressing neurons and that estrogens could enhance their activity.
Next, we analyzed itch-related neural activity using in vivo extracellular recordings from neurons in superficial laminae I and II (<100 μm from the surface; SI Appendix, Fig. S5) of the rat spinal dorsal horn (Fig. 4A). Spontaneous firing of dorsal horn neurons was relatively low in the OVX + b or OVX + E group (SI Appendix, Fig. S6). Neuronal responses elicited by an innocuous von Frey filament were significantly lower in the OVX + E compared with the OVX + b group (Fig. 4 B-D: Mann-Whitney U test: **P < 0.01, OVX + b versus OVX + E), consistent with the behavioral data (Fig. 1H). Subsequently, GRP was superfused to the surface of the spinal cord to search for responsive neurons. A total of 25% of neurons (14/55) in the OVX + b group and 30% (16/54) of neurons in the OVX + E group responded to GRP (SI Appendix, Fig. S7). These GRPresponsive neurons were then tested for responsiveness to intradermal hind paw injection of histamine. A total of 31% of neurons in the OVX + b group responded to histamine, while a significantly higher proportion, 75%, in the OVX + E group responded to histamine (χ 2 test: P < 0.01) (SI Appendix, Table  S1). The duration of histamine-evoked firing was significantly longer in the OVX + E group than in the OVX + b group (Fig. 4 E-H: Mann-Whitney U test: **P < 0.01; Mann-Whitney U test: *P < 0.05, OVX + b versus OVX + E). These results demonstrate that estrogen treatment increased the frequency and duration of histamine-evoked activity of spinal GRPRexpressing neurons, consistent with the behavioral data showing that estradiol increased the number of histamine-evoked scratch bouts and overall scratching duration ( Fig. 1 B, C, E, and F and SI Appendix, Fig. S1).
Finally, to understand the mechanism by which estrogens regulate itch transmission through GRP and GRPR, we focused on the nuclear estrogen receptor α (ERα), which is a transcription factor. After ligand binding, the nuclear receptors bind directly to the specific sequence of DNA on the target gene promotor known as the hormone responsive element to regulate transcription. We found the estrogen responsive element (ERE) on the promotor region of the genes for both GRP and GRPR. At first, to know the interaction of ER-α and GRPR, we analyzed the colocalization of ERα-and GRPR-positive neurons in the spinal dorsal horn. We found that many GRPR-expressing neurons coexpressed ER-α in the superficial layers of the spinal dorsal horn (Fig. 4 I-P: 85.7 ± 3.2% colocalization in male and 92.3 ± 1.5% in females). Next, the chromatin immunoprecipitation (ChIP) assay was used to examine whether the transcription factor ER-α binds to the GRP or GRPR promoter region and also to examine cAMP response element binding protein (CREB)-mediated transcription, because estrogens and ER-α have a CREB-mediated transcriptional pathway (20)(21)(22). In the spinal dorsal horn immunoprecipitation assay, we found that pCREB was bound to the GRP promoter ( Fig. 4 Q, Left) and ER-α and pCREB were bound to the GRPR promoter (Fig. 4  intact females. However, we did not observe ER-α or pCREB to bind to either the GRP or GRPR promotor in the OVX females ( Fig. 4 Q and R, Right). These results suggest that estrogens regulate the transcription of GRPR, which is essential for itch transmission, via ER-α and CREB, and that CREB is involved in the transcriptional regulation of GRP. Thus, estrogens selectively regulate histamine-evoked itch in females via the spinal GRP/GRPR system (Fig. 4S).
Several lines of evidence indicate that estrogens regulate histamine-evoked itch via activity of GRPR-expressing neurons in the spinal dorsal horn of female rats. First, physiological concentrations of estradiol replacement in ovariectomized female rats enhanced and sustained histamine-evoked scratching directed to the nape of the neck as well as biting behaviors directed to the hind paw. This enhancement was mediated via histamine H1 receptors and spinal GRPR-expressing neurons because an H1 receptor antagonist and a GRPR blocker inhibited the scratching in OVX + E rats. Second, estradiol replacement increased the histamine-evoked neural activity as assessed by c-Fos expression in GRP GRPR B D  spinal GRPR-expressing neurons. Third, electrophysiological data show that estrogens markedly increased the frequency and duration of histamine-evoked responses of spinal dorsal horn neurons without affecting spontaneous neural firing. Overall, our results directly demonstrate that the spinal GRP/GRPR system is specifically involved in female estrogen-dependent itch. In our study, physiological levels of estradiol treatment did not affect the threshold of thermal pain but decreased touch sensitivity and increased itch sensitivity via the itch-specific spinal GRP/GRPR system. Our results indicate that estradiol acts separately on the neural circuits that underlie these distinct somatosensory submodalities.
Estradiol not only increased itch sensitivity and activity of GRPR-expressing neurons but also the expression of GRP mRNA and the number of GRPR-positive neurons in the spinal F I  n n n n n n n n n n) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ( (s (s (s s (s (s (s (s (s s (s ( (s (s ( (s (s (s c c c c c c c c c c c c c c c c c c c ec e e e e ec ec ec ec e e e ec e ec e e e ec e ) ) ) ) ) ) ) ) ) ) ) ) ) ) )

S S S S S S S S S S S S S t t t t t t t t t h h h h h h h h hi i i i i i i i i S S S S S S S S S S
) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) n n n n n n n n n n n n n n n n n n n  ( ( ( ( (  ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) s ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) dorsal horn of female rats. In contrast, progesterone had no effect on itch sensitivity or GRP expression. Treatment with both estradiol and progesterone resulted in increased GRP expression but not itch behavior, suggesting that progesterone inhibits the effects of estradiol on itch transmission in females. In the present study, we found a high degree of colocalization of ER-α and GRPR in the spinal dorsal horn. We also showed that estradiol affects the transcription of GRP and GRPR in the spinal dorsal horn via the transcription factor, ER-α, binding to ERE and CRE (cAMP response element) on the GRP and GRPR promotor regions. Because the CRE site was reported to be necessary for transcriptional activity of the GRPR promotor (27)(28)(29), estrogens can interact with CREB through ER-α to regulate transcription of several genes (20)(21)(22). It has been recently reported that an ERαpositive subpopulation of interneurons is involved in itch transmission in the spinal dorsal horn (30) and that activation of ER-α is associated with inflammatory pruritus in mice (31). Thus, our data indicates that ER-α is involved in the transmission of itch, and we have newly clarified that estrogens modulate itch sensitivity by regulating spinal GRP and GRPR gene transcription.
Histamine is the primary peripheral itch mediator in humans but elicits little if any itch-related scratching behavior in rats   (24,32); instead, serotonin elicits robust scratching in this species (32)(33)(34). This was confirmed by the present results in naïve male and female rats. On the other hand, our present study showed that, in the presence of estrogens, there was an enhancement of histamine-evoked itch sensitivity in female rats. Therefore, this rat model can be a useful for elucidating the mechanism of histamine-dependent itch. Spinal GRPR neurons are reported to be required for histaminergic and nonhistaminergic itch in male mice (11,35,36). Our study additionally supports a role for spinal GRPR-expressing neurons in histaminergic itch transmission in female rats. Previous studies indicated that spinal neurons that express the neuromedin B receptor (NMBR) are required for itch, including histaminergic itch. NMBR-expressing neurons function upstream of GRPR-expressing neurons in the spinal dorsal horn (37). This presynaptic or neuromedin B (NMB)/NMBR-mediated mechanism might play an important role, since histamine activated a higher proportion and longer duration of firing of GRPresponsive neurons in the OVX + E group than in OVX + b group, despite the observation that comparable proportions of neurons in these two groups responded to exogenous superfusion of GRP (25% in the OVX + b and 30% in the OVX + E group).
The increased sensitivity to histamine-induced itch in the estrogen-treated group may be due to the transmission of itch via NMBR-expressing neurons. In addition, like histamine, estradiol enhanced itch sensitivity elicited by CQ, which also exhibited a female-dominant sex difference. A recent study demonstrated that GRP in sensory neurons is required for CQ-but not histamine-evoked itch behavior using GRP cre-KI mice and that spinal GRP neurons are dispensable for itch transmission (38). In contrast, our results showed that estrogen treatment increased GRP mRNA in the spinal cord, GRPR neural activity in the spinal dorsal horn, and itch behavior induced by histamine and CQ in female rats. Furthermore, estrogen treatment affected the transcription of GRP and GRPR in the spinal dorsal horn via ER-α and CREB, suggesting that responses of GRPR neurons to GRP are increased, with estrogens playing a regulatory role. In addition, genetic and pharmacological approaches provide evidence supporting the role of spinal GRP neurons in itch. Chemogenetic activation of GRP neurons in the spinal cord increased the CQ-and histamineinduced itch-related behavior but did not change the pain response (15). Chemogenetic inhibition of GRP neurons in the spinal cord suppressed both histaminergic and nonhistaminergic itch without affecting the mechanical pain threshold (39).
ER-α has been demonstrated to be expressed in smalldiameter neurons of rat dorsal root ganglia (DRG), and its expression is regulated by circulating levels of 17β-estradiol (40). Therefore, it is possible that estrogens may additionally regulate GRP and NMB expression in the DRG via ER-α to increase scratching behavior induced by histamine and CQ (Fig. 4S).
In addition to pruritus during the period of pregnancy, there is an increase in the prevalence of atopic dermatitis in female patients at the onset of adolescence, when the blood estrogen concentration rapidly increases (4,7,41). Certain itchy skin diseases and allergic conjunctivitis with symptoms of itching show a greater prevalence in females than males (42,43). Our findings may be related to these female-specific itchy conditions, but our study has certain limitations. First, the study was performed using an animal model, and there is no supporting evidence that a similar mechanism pertains to humans. Secondly, we used hormone replacement as a surrogate for pregnancy and did not investigate other changes during gestation that might contribute to itch. Third, while our data provides correlative evidence for the effects of estradiol on spinal GRP signaling, we did not currently perform loss-of-function experiments that are needed to demonstrate causality of the proposed mechanisms; these should be done in future studies. Taken together, we hope that our research will serve as the basis for the treatment of pruritus in women and a bridge to gender medicine.

Materials and Methods
All experimental procedures were approved in accordance with the Guide for the Care and Use of Laboratory Animals (44) prepared by Okayama University (Okayama, Japan), by Kyoto Prefectural University of Medicine (Kyoto, Japan), by Toyama University (Toyama, Japan), and by the National Institute of Genetics (Shizuoka, Japan) and performed in accordance with the NIH guidelines on animal care. Adult wild-type Wistar rats and GRPR-mRFP transgenic rats were used in this study. To examine the effects of female sex steroid hormones, ovariectomized females were implanted with a blank capsule, a capsule containing physiological levels of 17β-estradiol, a capsule containing physiological levels of progesterone, or a capsule containing physiological levels of both hormones, for 1 to 2 mo. For itch behavioral analysis, rats received either saline, 3% histamine, or 10% CQ diphosphate salt diluted in saline via intradermal injection in the nape of the neck, cheek, or hind paw. Immediately after the injection, the rat was placed into the arena and videotaped from above for 60 min for scratching behavior. Mechanical sensitivity was assessed by the von Frey filament test, and thermal pain sensitivity was assessed by the Hargreaves test. For RT-PCR, enzyme-linked immunosorbent assay, Western blot, and the ChIP assay, the dorsal horns of the cervical spinal cords were collected and used for analysis. Complete methods are described in SI Appendix. Brain and spinal cord sections were used for immunofluorescence and immunoperoxidase histochemistry after perfusion with physiological saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer. Antibody information is provided in SI Appendix, Fig. S2. In vivo extracellular single-unit recordings of superficial spinal dorsal horn (lamina I and II) neurons were performed in female rats. Statistical analyses were performed using SPSS Statistics version 27 (IBM). Graphs were made using GraphPad Prism 8 (GraphPad Software). Statistical analyses for each study are indicated in the figure legends. More detailed information on materials and methods is provided in SI Appendix.
Data Availability. All study data are included in the article and/or SI Appendix.