A neural circuit for sex-dependent conditioned pain hypersensitivity in mice

Zhang, Mingjun; Ni, Ziyun; Ma, Jun; Liu, An; Liu, Ying; Lou, Qianqian; Dong, Wan-Ying; Zhang, Zhi; Li, Juan; Cao, Peng

doi:10.1038/s41467-025-58851-x

Download PDF

Article
Open access
Published: 17 April 2025

A neural circuit for sex-dependent conditioned pain hypersensitivity in mice

Mingjun Zhang¹^na1,
Ziyun Ni¹^na1,
Jun Ma¹^na1,
An Liu²^na1,
Ying Liu¹,
Qianqian Lou¹,
Wan-Ying Dong¹,
Zhi Zhang ORCID: orcid.org/0000-0002-4205-3181^1,2,3,
Juan Li¹ &
…
Peng Cao ORCID: orcid.org/0009-0001-6712-429X^1,3

Nature Communications volume 16, Article number: 3639 (2025) Cite this article

11k Accesses
6 Citations
9 Altmetric
Metrics details

Subjects

Abstract

The neural mechanisms underlying sex-specific pain, in which males and females exhibit distinct responses to pain, remain poorly understood. Here we show that in a mouse model of male-specific pain hypersensitivity response to pain conditioning environments (contextual pain hypersensitivity model), elevated free-testosterone leads to hyperactivity of glutamatergic neurons in the medial preoptic area (Glu^mPOA) through activation of androgen receptor signaling, which in turn induces contextual pain hypersensitivity in male mice. Although not observed in naïve female mice, this pain phenotype could be induced in females via chronic administration of testosterone propionate. In addition, Glu^mPOA neurons send excitatory inputs to GABAergic neurons in the ventrolateral periaqueductal gray (GABA^vlPAG) that are required for contextual pain hypersensitivity. Our study thus demonstrates that testosterone/androgen receptor signaling enhances Glu^mPOA→ GABA^vlPAG pathway activity, which drives a male-specific contextual pain hypersensitivity, providing insight into the basis of sexually dimorphic pain response.

Divergent modulation of pain and anxiety by GABAergic neurons in the ventrolateral periaqueductal gray and dorsal raphe

Article Open access 16 December 2022

Modulation of trigeminal neuropathic pain by optogenetic inhibition of posterior hypothalamus in CCI-ION rat

Article Open access 10 January 2023

Dorsal raphe GABA-ergic neurons regulate the susceptibility to social transfer of pain in mice

Article 26 February 2025

Introduction

Males and females in humans and rodents have different responses to various nociceptive stimuli and endogenous pain^1,2. Although females are generally considered more susceptible to developing pain symptoms than males^3,4, previous studies have reported male-specific pain hypersensitivity^5,6. The mechanisms underlying these sexually dimorphic phenomena have remained largely unknown.

Sex hormones have been implicated in pain processing in mice^7,8,9. Female mice require significantly longer recovery times in neuropathic pain models, possibly due to enhanced nociceptor excitability stimulated by endogenous estrogen metabolites^10,11,12. In male mice, testosterone plays a crucial role in nociceptive systems^13,14. Castration surgery can significantly alleviate hypersensitivity in contextual pain or inflammatory pain model mice, while administration of testosterone propionate can restore the pain hypersensitivity^5,15. However, results to the contrary have also been reported in both clinical and preclinical studies^16,17. These findings suggest that sex hormones may have complex and multi-faceted roles in modulating pain, exerting pro- or anti-nociceptive effects in a manner dependent on the types of pain^18,19. It is still unknown how different brain structures and their corresponding pathways may be responsive to hormones, or how aberrant activity in these regions can result in sex-dependent conditioned pain hypersensitivity.

Using a combination of viral tracing, in vivo fiber photometry, multi-tetrode recordings, and optogenetic methods, we dissected the organization and explored the function of an mPOA → vlPAG pathway required for male-specific contextual pain hypersensitivity in mice. Furthermore, we investigated whether reproductive hormones were associated with the sexually dimorphic activation of this pathway using in vivo microdialysis-ELISA and pharmacological tools. Findings in this study reveal an important role of free-testosterone/androgen receptor signaling in enhancing mPOA excitatory projections to vlPAG^GABA in male-specific contextual pain hypersensitivity mice.

Results

Conditioned context induces hyperactivity of Glu^mPOA neurons in male mice

We initiated our study of the neural mechanisms underlying sex-specific pain hypersensitivity (i.e., pain sensitization) by establishing a murine model for conditioned contextual pain hypersensitivity^5,6. Very briefly, we induced abdominal constriction in male and female mice by intraperitoneal (i.p.) injection of acetic acid in a specific context, which induced the same writhing and constriction response in both sexes (Fig. 1a–c, and Supplementary Fig. 1a, b). We then monitored both mechanical and thermal pain thresholds the following day during exposure to the same or different environmental contexts as that of the acetic acid injection, by switching rooms or test cylinders (Fig. 1a). We observed that male mice indeed exhibited pain hypersensitivity upon exposure to the same context (male^SC), but not in different contexts (male^DC); whereas no contextual pain hypersensitivity was observed in females or saline-treated controls (Fig. 1d, e, and Supplementary Fig. 1c–g). This contextual pain hypersensitivity persisted for at least 5 days in male mice (Fig. 1f, g). In addition, no significant differences in pain hypersensitivity behaviors were observed between the day and night (Supplementary Fig. 1h, i), and no anxiety-related behaviors were observed in both male^SC and female^SC mice (Supplementary Fig. 2a–d). Furthermore, we found that male-specific contextual pain hypersensitivity could also be induced by the same auditory cues in the pain-associated environmental context (Fig. 1h–j).

**Fig. 1: Contextual pain hypersensitivity in response to pain conditioning in male, but not female, mice.**

To identify the potential brain regions activated during this male^SC pain hypersensitivity, a widely used screening approach was employed based on the detection of c-Fos—an immediate-early marker of neuronal activity—which identified 11 regions across the whole brain for evaluation. We found that the ratio of male^SC c-Fos⁺ cell counts to male^DC c-Fos⁺ cell counts was significantly altered in multiple brain regions compared to that of females at day 2 (D2) (Supplementary Fig. 3a–e). In addition, we detected the highest ratio of increased c-Fos⁺ cells in the mPOA, which is known to participate in hormone-responsive sexually dimorphic behaviors^20,21 (Supplementary Fig. 3f). Further immunofluorescence staining revealed colocalization of c-Fos with approximately 80% of mPOA cells positive for an antibody marking glutamatergic neurons, and 15% of c-Fos signal co-labeling neurons positive for GABAergic-specific antibody (Supplementary Fig. 3g). These findings support a likely functional role of Glu^mPOA neurons in the observed male^SC pain hypersensitivity phenotype.

Whole-cell patch-clamp recordings showed that the firing rate of Glu^mPOA neurons was significantly increased in male^SC compared to male^DC mice, while no such difference was observed between female^SC and female^DC mice at D2 (Supplementary Fig. 4a, b). Notably, this increased Glu^mPOA neuronal activity in male^SC mice persisted through D4, but was abolished by D8 (Supplementary Fig. 4c, d). To test whether this increase in Glu^mPOA neuronal activity could be attributed to changes arising from the establishment of conditioned context pain hypersensitivity, fiber photometry recordings were performed in freely moving VgluT2-Cre mice injected with an AAV expressing Cre-dependent GCaMP6m (AAV-DIO-GCaMP6m) in the mPOA (Fig. 2a, b). We found that the activity of GCaMP6m-expressing Glu^mPOA neurons was significantly increased in male^SC mice compared to their male^DC counterparts, with no apparent difference in the female groups at D2 (Fig. 2c–e); and no such increase was found in either Vgat-Cre male^SC or female^SC mice administered with mPOA injection of an AAV-DIO-GCaMP6m virus (Supplementary Fig. 5a–d). To exclude the impact of transferring mice by hand on Glu^mPOA neuronal activity, we transferred mice from one different context (plexiglass cylinder with horizontal stripe, DC1) to another (plexiglass cylinder with vertical stripe, DC2) as a negative control. We detected a slight, transient increase of calcium activity in Glu^mPOA neurons during the contextual shift, which immediately returned to the baseline level observed in the DC1 (Supplementary Fig. 5e–h). Subsequent in vivo multi-tetrode recordings of optogenetically tagged Glu^mPOA neurons in freely moving mice (Fig. 2f–h) showed that the spontaneous firing rate of Glu^mPOA neurons was significantly increased at 5 minutes, but later recovered to baseline levels within 3 h post acetic acid injection in both sexes (Supplementary Fig. 6). At D2, the spontaneous firing rate were significantly higher in male^SC than male^DC mice, but not in female^SC versus female^DC mice (Fig. 2i, j). This increased Glu^mPOA neuronal activity in male^SC mice also exhibited temporal kinetic changes (persisted through D4, but was abolished by D8) (Supplementary Fig. 7), which was significantly correlated with male-specific contextual pain hypersensitivity behaviors over time (Fig. 1f).

**Fig. 2: Conditioned context elicits pain hypersensitivity by activating Glu^mPOA neurons.**

To test whether Glu^mPOA neurons mediated pain hypersensitivity in male^SC mice, an AAV-DIO-ChR2-mCherry was infused into the mPOA of VgluT2-Cre mice (Fig. 2k, l). Pain thresholds assays and conditioned place preference tests showed that optical activation of Glu^mPOA neurons resulted in both pain hypersensitivity and place avoidance in both male^DC and female^DC mice (Fig. 2m, n, and Supplementary Fig. 8). By contrast, optical inhibition of Glu^mPOA neurons expressing AAV-DIO-eNpHR3.0-EYFP abolished contextual pain hypersensitivity of male^SC mice compared to control mice infected with AAV-DIO-EYFP virus (Fig. 2o–q); however, inhibition of the Glu^mPOA neurons had no effect on mechanical or thermal pain thresholds in female^SC mice at D2 (Fig. 2r). Further chemogenetic inhibition of Glu^mPOA neurons via infusion of the bilateral mPOA with an AAV-DIO-hM4Di-mCherry in VgluT2-Cre mice also resulted in abolishing male^SC pain hypersensitivity after i.p. injection of clozapine-N-oxide (CNO) at D2, but had no effect in control animals (Supplementary Fig. 9). These results collectively support a scenario in which pain conditioning could induce context-dependent hyperactivity in Glu^mPOA neurons only in male mice.

As other studies have shown that GABA^mPOA neurons can locally inhibit Glu^mPOA neurons²², we therefore tested whether GABA^mPOA neurons could modulate contextual pain hypersensitivity by injecting AAV-DIO-ChR2-mCherry virus into the mPOA of both male and female Vgat-Cre mice (Supplementary Fig. 10a). Behavioral assays indicated that optical activation of GABA^mPOA neurons could indeed reverse contextual pain hypersensitivity in male^SC mice compared with mCherry-infected control mice, but had no effect on female mice (Supplementary Fig. 10b, c). By contrast, optical inhibition of GABA^mPOA neurons resulted in pain hypersensitivity in both male^DC and female^DC mice (Supplementary Fig. 10d–f). These results indicate that GABA^mPOA neurons participated in conditioned contextual pain hypersensitivity.

Free-testosterone mediates contextual pain hypersensitivity in male mice

Since free-testosterone is the biologically active, unbound fraction of testosterone known to regulate sexually dimorphic behaviors through binding with androgen receptors in male mice²³, we hypothesized that free-testosterone could play a role in the development of contextual pain hypersensitivity. ELISA assays of free-testosterone in serum and plasma showed that male^SC mice had significantly higher levels than both male^DC and naïve male mice, and this elevation persisted through D4, returning to baseline levels by D8 (Fig. 3a and Supplementary Fig. 11a).

**Fig. 3: Free-testosterone contributes to male-specific contextual pain hypersensitivity.**

To examine the effects of artificially reducing free-testosterone on contextual pain hypersensitivity in male mice, we performed castration surgery and subsequently administered daily subcutaneous injections of the testosterone analogue, testosterone propionate, to restore free-testosterone in castrated male mice (Fig. 3b, c, and Supplementary Fig. 11b). At two weeks after surgery, nociceptive tests indicated that paw withdrawal threshold and latency were significantly increased in castrated male^SC mice compared with the sham male^SC group (Fig. 3d, e), which was accompanied with a significant decrease in Glu^mPOA neuronal activity (Fig. 3f, g). Conversely, the castrated male^SC mice treated with testosterone propionate had significantly lower paw withdrawal threshold and latency compared to the corresponding vehicle controls at D2 (Fig. 3d, e). Interestingly, nociceptive tests revealed that testosterone propionate-treated female^SC mice displayed a contextual pain hypersensitivity phenotype similar to male^SC mice, with lower thresholds than vehicle-treated female^SC controls, although no difference in paw withdrawal threshold or latency were detected between testosterone propionate- and vehicle-treated female mice before acetic acid injection (Fig. 3h–j).

To exclude the involvement of neuroprotective estrogen in contextual pain hypersensitivity, we also established the contextual pain hypersensitivity model in ovariectomized (OVX) female mice (Fig. 3k). Behavioral tests indicated that OVX female^SC mice still did not exhibit contextual pain hypersensitivity while daily subcutaneous injection of testosterone propionate for two weeks could induce contextual pain hypersensitivity in OVX female^SC mice (Fig. 3l). These results demonstrate that testosterone, rather than estrogen, mediates sex-specific contextual pain hypersensitivity.

It is well known that acute stress induces high levels of corticosterone, which confers analgesic effects, while chronic stress induces pain sensitization^24,25. To investigate whether stress contributes to male-specific contextual pain hypersensitivity, we subcutaneously injected male^SC mice with corticosterone synthesis inhibitor (Metyrapone) 30-min before testing at D2 (Fig. 4a), and found that both the conditioned contextual pain hypersensitivity and hyperacivity of Glu^mPOA neurons were reversed in Metyrapone-treated male^SC mice (Fig. 4b–d). In addition, serum corticosterone and free-testosterone levels were significantly reduced in the Metyrapone treatment group compared to saline-treated male^SC controls (Fig. 4e). Interestingly, we found that there were no changes in pain thresholds (Supplementary Fig. 11c, d) or in the serum free-testosterone concentration, but serum corticosterone levels were significantly elevated in castrated male^SC mice compared to the castrated male^DC group (Fig. 4f). These results demonstrate that free-testosterone, rather than corticosterone, is required for male-specific contextual pain hypersensitivity.

**Fig. 4: Stress and memory system regulate male-specific contextual allodynia.**

Given that different pain-associated contexts (i.e., spatial shapes, sound stimuli) could induce male-specific contextual pain hypersensitivity, we speculated that the memory system might trigger this response. To test this possibility, we delivered Zeta inhibitory peptide (ZIP), a PKMζ inhibitor, via bilaterally implanted cannulas into the dorsal hippocampus CA1 (dCA1) of male^SC mice on D2 to inhibit the formation of episodic memory (Fig. 4g). We found that the conditioned contextual pain hypersensitivity was abolished and that serum corticosterone and free-testosterone concentrations were significantly lower in ZIP-treated male^SC mice compared with vehicle-treated controls (Fig. 4h–j). Since direct inhibition of PKMζ using ZIP produces a wide range of effects^26,27, we also directly injected an AAV virus expressing tetanus neurotoxin (AAV-hSyn-TetTox-EGFP, TetTox) to inactivate synaptic transmission in the dCA1 of male mice (Fig. 4k). We found that both the conditioned contextual pain hypersensitivity and hyperacivity of Glu^mPOA neurons were reversed in TetTox-infected male^SC mice (Fig. 4l–q). These findings indicate that episodic memory was indeed responsible for triggering male-specific contextual pain hypersensitivity.

Testosterone induces contextual pain hypersensitivity through enhancing Glu^mPOA neuronal excitability

In vivo microdialysis in conjunction with ELISA indicated that free-testosterone contents were indeed significantly elevated in the mPOA of male^SC mice compared to male^DC and naïve male mice on D2 (Fig. 5a, b). To evaluate the effect of free-testosterone on Glu^mPOA neuronal activity, we then conducted whole-cell recordings in the mPOA brain slices incubated in 0.05 μM, 0.3 μM, and 1 μM testosterone propionate, respectively, which showed that firing rates were increased while the rheobase decreased in Glu^mPOA neurons incubated in both 0.3 μM and 1 μM testosterone propionate compared with the Glu^mPOA vehicle controls (Fig. 5c, d). Due to the potential effects on neuronal excitability arising from changes in synaptic transmission efficiency, we further examined spontaneous excitatory postsynaptic currents (sEPSCs) and spontaneous inhibitory postsynaptic currents (sIPSCs) of Glu^mPOA neurons after incubation with 1 μM testosterone propionate. Electrophysiological recordings showed that sIPSC frequency was significantly reduced in Glu^mPOA neurons, while no change was detected in the frequency or amplitude of sEPSCs (Supplementary Fig. 12a–d).

**Fig. 5: Effects of androgen receptor on contextual pain hypersensitivity.**

To monitor changes in Glu^mPOA neuronal activity in real-time in male contextual pain hypersensitivity model mice with artificially manipulated androgen levels, we acquired in vivo multi-tetrode recordings in castrated male mice with or without testosterone propionate. These electrophysiological recordings showed that Glu^mPOA firing rates were significantly decreased in castrated male^SC mice compared to the sham male^SC controls, but that this Glu^mPOA activity could be recovered by daily testosterone propionate injection in castrated male^SC mice compared to vehicle-injected controls (Fig. 5e, f). In addition, no significant difference in Glu^mPOA neuronal activity was observed in male^DC mice given the same treatments described above (Fig. 5e, f). These findings were confirmed by in vivo fiber photometry recordings in freely moving male mice, and Glu^mPOA neuronal activity was restored in castrated male^SC mice by testosterone propionate injection compared to vehicle-treated castrated male^SC mice (Fig. 5g–i). Moreover, microinfusion with testosterone propionate into the mPOA of naïve male mice decreased paw withdrawal threshold and latency compared to vehicle-treated control mice (Supplementary Fig. 12e–g). These cumulative findings suggest that elevated free-testosterone levels could elicit contextual pain hypersensitivity by enhancing Glu^mPOA neuronal excitability in male mice.

Immunofluorescence staining showed that the androgen receptor was significantly higher expression in the mPOA of male mice than that of female mice (Supplementary Fig. 13). We then locally infused an androgen receptor-specific antagonist, hydroxyflutamide, by implanted cannulas in the bilateral mPOA of male mice (Supplementary Fig. 14a). Behavioral tests and whole-cell patch-clamp recordings showed that both the contextual pain hypersensitivity and hyperexcitability of Glu^mPOA neurons were reversed in hydroxyflutamide-treated male^SC mice compared with the corresponding vehicle controls (Supplementary Fig. 14b–e).

We then constructed a Cre-dependent AAV virus vector expressing androgen receptor short-hairpin RNAs (shAR) to induce androgen receptor knockdown in Glu^mPOA neurons of VgluT2-Cre mice (Fig. 5j, k). Glu^mPOA specific androgen receptor knockdown resulted in significantly increasing the paw withdrawal threshold and latency in AAV-shAR-EGFP-infected male^SC mice, but not in AAV-EGFP-infected controls (Fig. 5l). In vivo multi-tetrode recordings showed that the spontaneous firing rate of Glu^mPOA neurons was significantly decreased in the androgen receptor knockdown male^SC mice compared with the AAV-EGFP control group at D2, but showed no difference in firing rates between the AAV-shAR-EGFP and AAV-EGFP male^DC groups (Fig. 5m, n). In addition, the paw withdrawal threshold and latency were significantly decreased in AAV-EGFP-infected naïve male mice with testosterone propionate infusion of the mPOA, but not in mice infected with AAV-shAR-EGFP (Fig. 5o, p). These results suggest that free-testosterone-induced contextual pain hypersensitivity in an androgen receptor-dependent manner in male mice.

An excitatory Glu^mPOA → GABA^vlPAG pathway

Although some studies suggest a potential role for the mPOA in pain, it has not been recognized as a classical component of the pain processing^28,29. Therefore, we hypothesized that the regulatory effects of Glu^mPOA neurons in contextual pain hypersensitivity might be routed through pain-related brain areas. To test this possibility, we performed anterograde tracing by injection of AAV-DIO-ChR2-mCherry virus into the mPOA of VgluT2-Cre mice to investigate multiple brain regions innervated by Glu^mPOA neurons (Fig. 6a). mCherry⁺ terminals originating from Glu^mPOA neurons were observed in the nucleus accumbens, bed nucleus of the stria terminalis, lateral hypothalamic area, lateral septum, ventral tegmental area, and ventrolateral periaqueductal gray (vlPAG) (Fig. 6b and Supplementary Fig. 15a). In particular, the vlPAG had the greatest abundance of mCherry⁺ signals (Supplementary Fig. 15b, c). By contrast, mCherry signals were scarce in classical pain processing nuclei, such as the anterior cingulate cortex (ACC) and primary somatosensory cortex (S1) (Supplementary Fig. 15b)^30,31, therefore suggesting the possibility that mPOA projections to vlPAG mediate contextual pain hypersensitivity. No significant differences of mCherry⁺ mPOA neuronal fibers in the vlPAG between male and female mice (Supplementary Fig. 16a–c). Further immunofluorescence staining showed that vlPAG had fewer androgen receptor-positive cells than mPOA in male mice, which partially excluded the possibility that free-testosterone could act directly on vlPAG to modulate pain (Supplementary Fig. 16d, e).

**Fig. 6: Effects of Glu^mPOA→ GABA^vlPAG pathway on contextual pain hypersensitivity.**

To examine the cell types in mPOA → vlPAG anatomical projections, we used an alternative anterograde monosynaptic tracing strategy in which AAV2/1-Cre-EGFP virus was injected into the mPOA, while an AAV-DIO-EGFP virus was injected into the vlPAG of C57 mice (Fig. 6c, d). Three weeks later, we observed EGFP⁺ neurons in the vlPAG that co-stained with either GABA-specific (~70%) or glutamate-specific (~20%) antibodies (Fig. 6e, f). In addition, retrograde trans-monosynaptic tracing by injection of a retroAAV-hSyn-EGFP into the vlPAG resulted in labeling EGFP⁺ neurons in multiple brain regions (Supplementary Fig. 17a–c). Among these regions, EGFP⁺ neurons in the mPOA were mainly co-labeled with glutamate antibodies (~70%) (Supplementary Fig. 17d, e). Moreover, after infusing a modified rabies virus (EnvA-pseudotyped RV-ΔG-EGFP) and Cre-dependent helper viruses (AAV-Ef1α-DIO-TVA-mCherry and AAV-Ef1α-DIO-RVG) in both male and female Vgat-Cre mice (Supplementary Fig. 18a), we found abundant EGFP⁺ neurons in the mPOA wherein they also colocalized with glutamate-specific antibody (~85%); no significant differences were observed in either the number of EGFP⁺ vlPAG-projecting mPOA neurons nor in the number of mCherry⁺-EGFP⁺ co-labeled starter cells (yellow) in the vlPAG between male and female mice (Supplementary Fig. 18b–h). These results suggest a Glu^mPOA→ GABA^vlPAG connection present in both male and female mice.

To characterize the function of Glu^mPOA → GABA^vlPAG connections, optogenetics experiments were performed (Fig. 6g). Using whole-cell recording in brain slices, brief light stimulation of ChR2-containing Glu^mPOA terminals in the vlPAG evoked reliable excitatory postsynaptic currents (EPSCs) from GABA^vlPAG neurons, and that treatment with AMPA receptor antagonist, 6.7-dinitroquinoxaline-2,3-dione (DNQX), could block these EPSCs (Fig. 6h, i). These results demonstrate that GABA^vlPAG neurons receive functional monosynaptic excitatory glutamatergic projections from Glu^mPOA neurons.

Essential role of the Glu^mPOA→ GABA^vlPAG pathway in male contextual pain hypersensitivity

To assess the role of GABA^vlPAG neurons in contextual pain hypersensitivity, we first conducted fiber photometry recordings in both male and female Vgat-Cre mice following vlPAG infusion with an AAV-DIO-GCaMP6m virus (Supplementary Fig. 19a). In moving male^DC mice into the same context used for acetic acid-induced pain conditioning, the calcium activity of GCaMP6m-expressing GABA^vlPAG neurons significantly increased, while female^DC mice displayed no such increase upon exposure to the pain-associated environmental context (Supplementary Fig. 19b–d). In addition, we infused either AAV-DIO-ChR2-mCherry or AAV-DIO-mCherry control virus into the vlPAG of Vgat-Cre mice (Supplementary Fig. 19e). In both male^DC and female^DC mice infected with AAV-DIO-ChR2-mCherry, optical activation of GABA^vlPAG neurons resulted in significantly decreasing paw withdrawal threshold and latency compared to AAV-DIO-mCherry-infected controls (Supplementary Fig. 19f, g).

To verify that the increased activity of GABA^vlPAG neurons was due to enhanced excitatory input from Glu^mPOA neurons in contextual pain hypersensitivity model mice, we infused AAV-DIO-hM4Di-mCherry into the mPOA and retroAAV-Cre into the ipsilateral vlPAG of C57 mice (Fig. 6j). Subsequent chemogenetic inhibition of vlPAG-projecting Glu^mPOA neurons resulted in increased paw withdrawal threshold and latency in hM4Di-expressing male^SC mice (Fig. 6k, l). In addition, activation of Glu^mPOA terminals in the vlPAG could induce pain hypersensitivity and place avoidance in both male^DC and female^DC mice (Fig. 6m, n, and Supplementary Fig. 20a–e). Conversly, inhibition of the Glu^mPOA→ GABA^vlPAG pathway resulted in alleviating contextual pain hypersensitivity in male^SC mice (Fig. 6o, p). Notably, as female^SC mice do not exhibit contextual pain hypersensitivity, optical inhibition of this pathway had no significant effects on their nociceptive thresholds (Supplementary Fig. 20f, g). These results thus demonstrate that the Glu^mPOA → GABA^vlPAG pathway plays a causal functional role in the occurrence of contextual pain hypersensitivity.

Discussion

The neural circuit mechanisms responsible for processing sex-specific pain has remained unknown. Here, we found a neural circuit from the hypothalamus to the midbrain mediates context-dependent pain hypersensitivity in male, but not female, mice, which requires activation of testosterone signaling to increase the activity of a Glu^mPOA → GABA^vlPAG circuit (Supplementary Fig. 21).

Numerous studies have shown that men and women differ in their responses to pain, with women commonly exhibiting increased pain sensitivity and a higher risk for chronic pain^1,2. However, emerging evidence suggests that genotype and endogenous opioid functioning play roles in these disparities^32,33. Clinical data show that men consume greater amounts of morphine postoperatively than women³⁴. Additionally, after surgery, women more frequently use sedative medication for pain relief, while men more frequently use analgesic medication³⁵. It is important to note that the contextual pain hypersensitivity phenomenon observed in mice in our study has also been demonstrated in humans. Specifically, male participants re-exposed to a context previously associated with the ischemic tourniquet test exhibit pain hypersensitivity, while female participants do not⁵. This clinical evidence implicates that in certain contexts, men may exhibit more sensitive pain responses than women.

The mPOA mediates a variety of sexually dimorphic functions through the expression of abundant sterol and steroid receptors^{36,37,38,39,40}. It bears mention that no pathway from the mPOA has yet been identified in modulating a response to pain. We found that the elevated free-testosterone elicits hyperactivity in the temporal kinetics of Glu^mPOA neurons that are correlated with male-specific contextual pain hypersensitivity in a pain conditioning environment, which leads to the activation of GABA^vlPAG neurons through a functional connection with the vlPAG—a key structure in descending pain modulation⁴¹. The neural pathway from the vlPAG sends excitatory projections to the rostral medial medulla which then project to the dorsal horn of the spinal cord, where the pathway is then available to modulate afferent nociceptive information^41,42. Additionally, GABA^vlPAG neurons play a central role in endogenous opioid-induced analgesia^43,44,45. Opioids act on the opioid receptors expressed on GABA^vlPAG neurons to inhibit GABA release, which in turn exerts anti-nociceptive effects^46,47. Aligning well with this understanding, findings in the current study show that optical activation of GABA^vlPAG neurons confers a pro-nociceptive effect in both sexes. Thus, the mPOA can serve as a bridge towards sexually dimorphic behaviors by responding to reproductive hormones via functional connectivity with multiple brain regions.

Acute stress is well-documented to commonly produce analgesic effects, while chronic stress induces pain hypersensitivity^48,49. Given the short period of time (1 day) during which contextual pain hypersensitivity was generated in the present study, it is unlikely that the observed pain sensitization could be attributed to chronic stress. We therefore speculate that the male-specific contextual pain hypersensitivity observed in our experiments was unlikely to reflect a chronic stress-associated increase in corticosterone (as reported by Martin et al.⁵); rather, our findings could be reasonably interpreted to illustrate a process in which inhibition of HPA axis activation reverses contextual pain hypersensitivity, accompanied by a reduction in corticosterone and free-testosterone levels. Supporting this scenario, castrated male mice exhibited a similar rescue of contextual pain hypersensitivity behaviors and retained elevated free-testosterone and serum corticosterone levels.

Previous reports have shown males can more effectively recall the stress-inducing properties of a conditioned context^50,51. For example, pharmacological inhibition or genetic blockade of the formation of synaptic plasticity required for episodic memory could decrease formalin-induced pain and referred visceral or muscle pain in male, but not female, rats⁵. Our results were consistent with this report and revealed that blocking episodic memory could reverse the elevation in free-testosterone levels and hyperexcitability of Glu^mPOA neurons. Thus, the present study supports a model in which: (i) episodic memory is first initiated in pain conditioning environments; (ii) this contextual memory leads to hyperactivation of the hypothalamic-pituitary-gonadal axis in mice; (iii) this hyperactivity increases serum free-testosterone levels; (iv) in turn, elevated free-testosterone enhances activity of the Glu^mPOA → GABA^vlPAG circuit by directly acting on androgen receptors on Glu^mPOA neurons; (v) Glu^mPOA → GABA^vlPAG activation ultimately mediates the observed contextual pain hypersensitivity in male mice.

Both female humans and rodents are more sensitive to pain than males^1,2, which may be due to sex-specific reproductive hormones, especially estrogen^52,53,54,55. Our study revealed that elevated free-testosterone leads to hyperactivity of Glu^mPOA neurons, which induces contextual pain hypersensitivity in male mice. Although did not occur in female mice, this pain phenotype could be induced by injection of testosterone propionate. These results suggest that free-testosterone is one of the potential substrates responsible for the difference in nociceptive perception between males and females, at least mediating male-specific contextual pain hypersensitivity that is absent in females. This understanding could guide future efforts focused on analogous, or entirely distinct, mechanisms in females and can certainly facilitate the experimental design of studies seeking to ascertain how distinct hormonal contexts influence pain sensitivity. Additionally, understanding the role of sex hormones, like testosterone, in modulating pain can open avenues for translational research. For example, it remains unknown whether similar pathways are modulated by estrogen or other hormones in females, which may contribute to the observed differences in the prevalence of chronic pain between sexes. Finally, our discoveries provide empirical support for the need to consider sex-specific factors in the development of pain management therapies, and may drive the discovery of more personalized and effective treatment strategies.

Overall, our identification of the mPOA as a fundamental brain region for mediating androgenic enhancement of nociceptive transmission opens avenues for research to expand understanding of the neural mechanisms responsible for sexually dimorphic responses to pain.

Methods

Animals

The experiments used adult (8–12 weeks) C57BL/6J (Charles River), VgluT2-Cre (Jackson Labs, 028863), Vgat-Cre (Jackson Labs, 028862), and Ai14 (RCL-tdT) (Jackson Labs, 007914) lines of both sexes. The VgluT2-Cre and Vgat-Cre were crossed to Ai14 mice to obtain the double heterozygous mice used in this study. The mice were group-housed with five per cage in a colony in a stable environment (23–25 °C ambient temperature and 50% humidity) unless a cannula or tetrode array was implanted. They were maintained under a 12-h light–dark cycle (lights on from 7:00 to 19:00) with ad libitum access to water and food. To generate litters, two females were mated with one male in our facility. The male was removed after 10 days and the females separated into individual cages 1–4 days prior to giving birth. All mice feeding and experiments were conducted in accordance with the requirements of the Animal Care and Use Committee of the University of Science and Technology of China (USTCACUC25080124017).

Pain model

The baseline sensitivity of both male and female mice to mechanical and thermal nociception was assessed by von Frey test and Hargreaves test three times at 5-min intervals before formal modelling (Day 0). On the first day of testing (conditioning day, Day 1), mice were injected intraperitoneally with saline or 0.9% acetic acid (AA) in a volume of 10 mL/kg⁵ and immediately placed into the conditioning context (plexiglas cubes within the testing room) for 1 h, after which they were placed back in their homecage and returned to the rearing room. On the second day (test day, Day 2), some mice were returned to the same cubes in the same room (same context, SC), and assessed their pain thresholds as baseline on Day 0. Other mice were placed within plexiglas cylinders instead of cubes, in a different testing room (different context, DC).

Castration, ovariectomy and testosterone supplement

Castration, ovariectomy and sham-surgery were performed on female mice and nociceptive tests were conducted at two weeks after recovery. In a separate experiment, castrated male and ovariectomy female mice were subcutaneously injected with testosterone propionate (Solarbio, IT0120), dissolved in DMSO (1 mg/mL), for 14 days at a dose of 1 mg/kg daily^56,57,58. Then, the mice were subjected to contextual pain hypersensitivity modelling and measured pain thresholds.

Behavioral paradigms

All behavioral tests were performed during the light phase (7:00 to 19:00). Operators were blinded to the experimental group during the scoring. On the first day (D1), mice were placed in the conditioning context, which consisted of rectangular plexiglas cubicles within a particular lab testing room. After adaption for 30 min, the mice were gently removed from their testing cubicles, injected with 0.9% acetic acid, and immediately returned to the cubicles for 1 h to pair with context cues. Afterward, they were placed back in their homecage. The next day (D2), some mice were returned to the same cubicles in the same room (same context, SC) for 30 min and assessed for behavior (i.e., pain threshold tests, open field tests, and elevated plus maze). Other mice were tested in a different context (DC), by being placed within plexiglas cylinders instead of cubicles, in a different lab testing room.

For auditory cues, mice in the same sound stimulation (SC, white noise sound, 65 dB) were compared to mice in a different intensity of sound (DC, ambient sound, 50 dB) on D2.

von Frey test

Mice were habituated in transparent plastic chambers positioned on a wire mesh grid for at least 3 days (1 h each day) prior to testing to minimize stress. The paw withdrawal threshold was assessed by using a series of calibrated von Frey filaments, which were perpendicularly applied to the middle of the plantar surface of the left hind paws with sufficient force to bend the filaments. The pressure was gradually increased, and a positive response was considered when the mice briskly presented withdrew or licked their paws. If there was no positive pain response, a filament with a greater force was applied, and the paw withdrawal threshold was calculated as the average of three applications. In optogenetic experiments, pain thresholds were tested before light stimulation and 30-min later when stimulated with blue or yellow light for 1-min in the contra-lateral hind paws.

Hargreaves test

The paw withdrawal latency was measured by using the Hargreaves apparatus (IITC Life Science Inc., USA). Mice were placed in the same manner as von Frey test for habituation. During the test, a radiant heat beam was focused on the left hind paw until a positive response happened, including the mice briskly presented a paw withdrawal, flinching, or licking. The time was recorded to determine the paw withdrawal latency and each mouse was measured at least three times to obtain an average threshold.

Open field test (OFT) and elevated plus maze (EPM)

For the OFT, a white plexiglas chamber (50 × 50 × 60 cm³) with a square central area (25 cm × 25 cm) was used to assay the anxiety-like behaviors. Mice were gently placed in the center, and their movements were recorded by a camera for 6 min. The entries into and time spent in the center of the open field were counted.

For the EPM, the apparatus consisted of a central platform (6 × 6 cm), two closed arms (30 × 6 × 20 cm), and two open arms (30 × 6 cm). The maze was elevated 100 cm above the ground. Mice were placed in the central area of the maze, with their heads facing the open arms, and allowed to explore freely. Movement trajectories were analyzed offline, and the time spent in and the number of entries into the open arms were calculated after recording.

The chamber was cleaned with 75% ethanol and clean water after each test to remove olfactory cues. The entries into and time spent in the center of the open field or the open arms were analyzed by EthoVision XT 14 software (Noldus, Wageningen, the Netherlands).

Conditioned place preference (CPP)

The experimental apparatus consisted of three compartments (12.7 cm × 12.7 cm) with a guillotine door separating the compartments. Mice could move freely in each compartment, and a video camera was hung on the top to record the movement of the mice. The first 15-min of recording was used as a baseline, following 15-min was used to assess place preference. To detect the preference of mice when Glu^mPOA neurons were activated, AAV-DIO-ChR2-mCherry was injected into the mPOA of VgluT2-Cre mice. After mild anesthesia with isoflurane, a fiber optic (Fiber Core = 200 μm, NA = 0.37) connected with a laser generator, and the mice were gently placed in the middle area and allowed to move freely. The stimulation protocol was the same as the optogenetic experiments. No stimulation was applied during the first 15-min recording. For the last 15 min, the 473-nm light was turned on once the mice were in the right area of the apparatus. The time spent in the right chamber were recorded during conditioned photostimulation of Glu^mPOA neurons for both male and female mice, along with the total distance traveled during the entire recording period (30 min).

Immunohistochemistry, imaging and image Analysis

Mice were deeply anesthetized with isoflurane (~15 s) and then perfused with saline for 3 min and 4% ice-cold paraformaldehyde solution for 4 min. After perfusion, the mouse brain was taken out carefully, and immersed in 4% paraformaldehyde solution overnight, then followed with 20% and 30% sucrose solution at 4 °C for dehydration until isotonic. Brains were cut into 40 μm coronal slices at −20 °C using a cryostat microtome system (CM1860, Leica). For immunofluorescence, the slices were washed 3 times with PBS. After being blocked in the buffer (containing 0.5% Triton X-100, 5% donkey serum in PBS) for 1 h at room temperature, they were incubated in primary antibodies, including anti-c-Fos (1:500, pig, Cat#226308, Synaptic Systems), anti-glutamate (1:500, rabbit, Cat# G6642, Sigma), anti-GABA (1:500, rabbit, Cat# A2052, Sigma), and anti-androgen receptor (1:500, rabbit, Cat#GB11253-100, Sigma) at 4 °C for 24 h. Then, slices were washed 3 times with PBS, and incubated with the secondary antibodies at room temperature in a dark place for 1.5 h. Cell nuclei were stained with DAPI (1:2000, Cat. No. D9542, Sigma) for 5 min. The fluorescence signals were imaged using an Olympus FV3000 microscope. Further analyses such as analysis of cell counts and co-localization were conducted using ImageJ software (Fiji edition, NIH) by an observer blind to the condition. For c-Fos immunohistochemistry, the contextual pain hypersensitivity model mice were placed in the different or same context for 30 minutes on D2.

Stereotaxic surgery

After being anesthetized with pentobarbital (20 mg/kg, i.p.), the mice were immobilized on the stereotaxic frame (RWD Life Science Inc., China). Sterile ointment was applied to each eye and a miniature heating pad was placed under the mice’s body to maintain the temperature at 37 °C. After disinfection and a midline scalp incision, the skull surface was exposed with a midline scalp incision and the position was leveled. A syringe tip and adjustable speed dental drill (B67275, Meisinger, Germany) was used to open a small (~0.5 mm) craniotomy. Then, the virus was injected into the brain region specifically at a speed of 30 nL/min by using a 10 μL microsyringe (Gaoge) with calibrated glass microelectrode (1B 100-3, WPI, USA). Following the injection, the microsyringe was left in the injection site for 5–10 min to minimize virus leakage in the track. Finally, the incision was sutured, and the surgical wound was sterilized. The mPOA coordinates relative to bregma were as follows: anterior/posterior (AP), 0.28 mm; medial/lateral (ML), 0.32 mm; dorsal/ventral (DV), −5.17 mm. The vlPAG coordinates: AP, −4.6 mm; ML, 0.45 mm; DV, −2.65 mm. The dCA1 coordinates: AP, 1 mm; ML, −1.94 mm; DV, −1.2 mm.

For anterograde tracing, rAAV-Ef1α-DIO-hChR2(H134R)-mCherry-WPRE-pA (AAV-DIO-ChR2-mCherry, AAV2/5, 3.63 × 10¹² vg/mL, 200 nL, BrainVTA) was injected into the mPOA of VgluT2-Cre mice. After three weeks, brain slices were prepared for examining the mCherry⁺ fibers originating from Glu^mPOA neurons in the whole brain. To visualize the vlPAG neurons innervated by the mPOA, the anterograde trans-synaptic virus rAAV-hSyn-CRE-EGFP-WPRE-hGH pA (AAV-hSyn-Cre-EGFP, AAV2/9, 5.60 × 10¹² vg/mL, 200 nL, BrainVTA) that can spread anterogradely to the downstream soma to express Cre was injected into the mPOA of C57 mice. Simultaneously, rAAV-EF1α-DIO-EGFP-WPRE-hGH-pA (AAV-DIO-EGFP, AAV2/5, 2.59 × 10¹² vg/mL, 100 nL, BrainVTA) was injected into the vlPAG. Three weeks later, the GFP signals and the co-staining with glutamate-specific or GABA-specific antibodies were investigated in the vlPAG.

For retrograde monosynaptic tracing, rAAV-hSyn-EGFP-WPRE-hGH pA (Retro-AAV-hSyn-EGFP, AAV2/R, 5.31 × 10¹² vg/mL, 100 nL, BrainVTA) virus was injected into the vlPAG, which could be absorbed by the terminals at the injection site and transported retrogradely to the soma to express the EGFP. At 3 weeks after the viral expression, mice were perfused transcardially, and brain slices were stained with a glutamate-specific or GABA-specific antibody in the mPOA. In addition, the helper viruses containing rAAV-EF1α-DIO-RVG-WPRE-hGH pA (AAV-DIO-RVG, AAV2/9, 4.59 × 10¹² vg/mL, BrainVTA) and rAAV-EF1α-DIO-H2B-mCherry-T2A-TVA-WPRE-hGH pA (AAV-DIO-TVA-mCherry, AAV2/9, 5.56 × 10¹² vg/mL, BrainVTA) (1:2, 150 nL) were injected into the vlPAG of Vgat-Cre mice. Three weeks later, RV-EnvA-ΔG-EGFP (3.10 × 10⁸ IFU/mL, 100 nL, BrainVTA) was injected into the same site of the vlPAG using identical conditions and coordinates. Seven days after the last injection, EGFP signals were stained with a glutamate-specific or GABA-specific antibody in the mPOA of the brain slices.

For optogenetic manipulation, the Cre-dependent virus rAAV-Ef1α-DIO-eNpHR3.0-EYFP-WPRE-pA (AAV-DIO-eNpHR3.0-EYFP, AAV2/9, 5.77 × 10¹² vg/mL, 200 nL, Braincase) or rAAV-Ef1α-DIO-hChR2(H134R)-mCherry-WPRE-pA (AAV-DIO-ChR2-mCherry, AAV2/5, 3.63 × 10¹² vg/mL, 200 nL, BrainVTA) was delivered into the mPOA in the right hemisphere of VgluT2-Cre mice to inhibit or active the Glu^mPOA neurons. For optogenetic activation of the Glu^mPOA → GABA^vlPAG pathway, the Cre-dependent virus, AAV-DIO-ChR2-mCherry was delivered into the mPOA and placed an optical fiber directly above the vlPAG of VgluT2-Cre mice. The mice injected with the Cre-dependent rAAV-EF1α-DIO-EYFP-WPRE-pA (AAV-DIO-EYFP, AAV2/9, 3.42 × 10¹² vg/mL, BrainCase) and rAAV-EF1α-DIO-mCherry-WPRE-hGH pA (AAV-DIO-mCherry, AAV2/9, 5.31 × 10¹² vg/mL, BrainCase) viruses at the same volume were used as controls.

For chemogenetic manipulation, the rAAV-DIO-hM4Di-mCherry-WPREs-pA (AAV-DIO-hM4Di-mCherry, AAV2/9, 5.29 × 10¹² vg/mL, 200 nL, BrainCase) virus was infused into the mPOA of VgluT2-Cre mice. For chemogenetic manipulation of the mPOA → vlPAG pathway, rAAV-hSyn-Cre-WPRE-hGH-pA (retroAAV-Cre, AAV2/R, 5.27 × 10¹² vg/mL, 100 nL, BrainVTA) was injected into the vlPAG and rAAV-DIO-hM4Di-mCherry-WPREs-pA (AAV-DIO-hM4Di-mCherry, AAV2/9, 5.29 × 10¹² vg/mL, 200 nL, BrainCase) virus was simultaneously infused into the mPOA of C57 mice. Three weeks after the viral injection, the contextual pain hypersensitivity mouse model was established. The mice injected with rAAV-EF1a-DIO-mCherry-WPRE-hGH pA (AAV-DIO-mCherry, AAV2/9, 5.45 × 10¹² vg/mL, BrainCase) virus into the mPOA as controls.

For synaptic inactivation of dCA1 neurons, AAV-hSyn-DIO-TetTox-EGFP (AAV2/9, 4.91 × 10¹² vg/mL, 100 nL, BrainVTA) was injected bilaterally into the dCA1. Three weeks after the viral injection, the contextual pain hypersensitivity mouse model was established.

Optogenetic and chemogenetic manipulation

After three weeks of viral expression, the mice were transported to a testing room and were habituated for approximately 4 h. Then, the mice were anesthetized with isoflurane for connecting the implanted optic fiber to a laser generator using a sleeve, and then, the mice were returned to the home cage for at least 30 min for adaption. Next, optogenetic stimulation experiments were carried out using a pulse of blue light (473 nm, 5–8 mW, 15 ms pulses, 20 Hz) or yellow light (594 nm, 5–8 mW, constant), controlled by a Master-8 pulse stimulator (A.M.P.I). For chemogenetic manipulation, the chemical ligand CNO (5 mg/kg, Sigma) was intraperitoneally injected in the mice under isoflurane anesthesia. Behavioral tests were then carried out at least 30-min later. For the control group, the same stimulus protocol was performed. After all behavioral tests were completed, the mice were sacrificed to verify the site of the viral injection and the optical fiber implantation. The data obtained from mice that missed the target brain regions were excluded from our analysis.

In vivo pharmacological approach

Guide catheters (diameter of 250 μm, RWD) were initially implanted into the mPOA or dCA1 of an anesthetized mouse bilaterally that were immobilized in a stereotaxic apparatus, and the cannula was secured to the skull with dental cement. After 2 weeks recovery, testosterone propionate (TP) and hydroxyflutamide (HF) were used to activate or antagonize androgen receptors (AR). The TP (0.3 μM, 300 nL) or vehicle (DMSO mixed with ACSF) was infused at a rate of 200 nL/min into the mPOA bilaterally of male mice⁵⁹. For inhibition of AR, HF (0.3 mM, 200 nL) or vehicle was administered at the same rate into the mPOA bilaterally of male contextual pain hypersensitivity C57 mice⁶⁰. For inhibition of αPKC, ZIP (Myr-Ser-Ile-Tyr-Arg-Arg-Gly-Ala-Arg-Arg-Trp-Arg-Lys-Leu, MedChem Express, HY-P1284) (10 nmol, 200 nL) or ACSF was administered at the same volume into the mPOA bilaterally of male contextual pain hypersensitivity C57 mice⁵. For inhibition of activation of the hypothalamic-pituitary-adrenal (HPA) axis, Metyrapone (2-methyl-1,2-di-3-pyridyl-1-propanone, MedChem Express, Su-4885) or saline was injected subcutaneously at a volume of 10 mL/kg (5 mg/mL), 30 min prior to testing on D2⁵.

Serum and plasma ELISA

The acetic acid-treated mice were placed in the same or different context for 30 min on D2. The mice were then deeply anesthetized with isoflurane, and blood was collected by rapidly removing the eyeballs, followed by decapitation. The blood was collected in heparin-infiltrated tubes for plasma. The samples were centrifuged at 4 °C in a high-speed centrifuge for 10 min to isolate serum or plasma for subsequent ELISA experiments.

In vivo microdialysis-ELISA

In vivo microdialysis in conjunction with enzyme-linked immunosorbent assay (ELISA) analysis was used to measure the free-testosterone concentration in the mPOA of freely moving mice. A microdialysis probe (CMA7, CMA) connected to a syringe infusion pump (CMA402, CMA) via polyethylene tube was initially implanted into the right mPOA of deeply anesthetized mice. The tissue was perfused with standard ACSF (for details in the electrophysiological recording section) via the pump at a rate of 0.25 μL/min and the dialysate was collected for 2 h through the probe in freely moving mice. The dialysate was frozen at −80 °C and the ELISA kits (Cat# D751047, Sangon biotech) or (Cat# D721183, Sangon biotech) were used to quantify the conceration of free-testosterone or corticosterone in the samples, according to the manufacturer’s instructions.

Western blot

The mPOA tissue was quickly obtained from 300 μm-thick slices cutting on the vibratome and homogenized in RIPA buffer, which contained a protease inhibitor cocktail, and 0.5% sodium deoxycholate. Then, the proteins were obtained by centrifuging at 12,000g at 4 °C for 15 min, and the protein concentrations were measured by BCA assay. Equal amounts of proteins were loaded onto 10% SDS-PAGE gels. Proteins were transferred to polyvinylidene difluoride (PVDF) membranes and blocked with 5% skim milk at room temperature for 1 h. The membrane was incubated with the primary antibody: anti-androgen receptor (1:500, Servicebio, GB115424-100) and β-actin (1:1000, Absin, abs137975) overnight at 4 °C, and with peroxidase-labeled secondary antibody (1:5000, Jackson, 115-035-003) at room temperature for 1.5 h. The protein bands were detected using high-sensitivity ECL reagent (GE Healthcare) and quantified by ImageJ software. Unprocessed scans of the blots are provided in the Source Data file.

Fiber photometry recording

To detect the calcium signals of Glu^mPOA and GABA^mPOA neurons, the AAV-DIO-GCaMP6m (AAV2/9, 6.09 × 10¹² vg/mL, 200 nL, Braincase) virus was injected into the mPOA, and an optic fiber was finally implanted above the mPOA of VgluT2-Cre or Vgat-Cre mice. For calcium signals of GABA^vlPAG neuronal soma recording, the same experimental manipulations were carried out in the vlPAG of Vgat-Cre mice. The fiber photometry recording was conducted at three weeks after viral expression, a fiber optic patch cord (Inper, MFO-1×2-F-W1.25-200-0.37-100) connected to the fiber photometry system (Inper) was attached to the implanted optic fiber using a ceramic sleeve with black heat-shrinkable tubes. To record fluorescence signals from GCaMP6m, light from a 470-nm LED was bandpass filtered (470/10 nm), collimated, reflected by dichroic mirrors (MD498, Thorlabs), coupled to an optical commutator (Doris Lenses) after being focused by a 20× objective lens (0.4 NA, Olympus), and then delivered at a power of 25–40 μW at the tip of the fiber optic cannula to excite GCaMP6m fluorescence. Then, the emitted fluorescence from GCaMP6m was bandpass filtered (525/40 nm, Thorlabs) and focused on the sensor of a CMOS camera. The end of the fiber was imaged at a frame rate of 40 fps with InperSignal, and the mean value of the ROI at the end-face of the fiber was calculated using InperPlot software. To serve as an isosbestic control channel, 410-nm LED light was bandpass filtered (410/10 nm) and delivered alternately with 470-nm LED light. GCaMP6m fluorescence intensity was then recorded during the conditioned context. The values of fluorescence change (ΔF/F) were derived by calculating ΔF/F (%) which were calculated as (F_duration − F_baseline)/F_baseline × 100%, where F_baseline is the mean of the GCaMP6m signals for 60 s to −5s in the different context and F_duration is the mean of GCaMP6m signals in the same context. All heatmaps and averaged calcium traces with shaded areas denoting the SEM were generated in InperPlot software (Inper Technology, Hangzhou). Each trail in the heatmap representing one mouse. To avoid interference in calcium signals, we handled the mice daily for at least three days to ensure that simple touch did not cause avoidance reactions. During the experiment, we maintained noise, light, and temperature within standard ranges. After placing the test mice in the DC (a plexiglass cylinder with open top and bottom) for 30 minutes, we gently and quickly transferred them by hand to a nearby SC chamber (a plexiglass cube with an openable top), with continuous recordings throughout the experimental period. For negative control, we transferred mice from a plexiglass cylinder with horizontal stripe (DC1) to a plexiglass cylinder with vertical stripe (DC2). The mice with missed targets of the optical fiber or failed viral injection would be excluded after the examination.

In vivo electrophysiology

Four homemade movable tetrode arrays, equipped with screw-driven microdrives, were employed to simultaneously record multichannel electrical signals from multiple neurons. Each tetrode, comprised of four twisted fine platinum/iridium wires (12.5-μm diameter, California Fine Wire, Grover Beach, CA), was implanted into the mPOA after 3 days where mice were housed in a single cage for environmental adaptation. The tetrodes were firmly affixed to the the skull using four miniature skull screws, cyanoacrylate glue, and dental cement.

Mice were allowed to rest for at least 7 days to recover and were habituated to the cables connected to the electrodes on their heads before recording. On D1, mice were administered 0.9% acetic acid following the same protocol as previously described. Subsequently, on D2, mice were placed to the same cubes within the same room (same context) or plexiglas cylinders in a different testing room (different context) to record data. Neurostudio software (Neurostudio, China) was utilized to record and process spiking activities, employing amplification, digitization at 40 kHz, and filtering within a 300–5000 Hz bandwidth. Then, the sorting of single unit spikes was conducted using Offline Spike Sorter software (Plexon Inc.) and Neuroexplorer 5 (Nex Technologies). The waveforms were first separated into individual clusters by Principal component analysis (PCA) and an unsupervised clustering algorithm based on a κ-means method. Data analysis only included well-isolated units (L ratio < 0.2, isolation distance > 15 and exhibit recognizable refractory periods (>2 ms) in the inter-spike interval (ISI) histograms).

For in vivo optogenetic tagging of Glu^mPOA neurons, VgluT2-Cre mice were unilaterally injected with AAV-DIO-ChR2-mCherry into the mPOA. Three weeks later, optrodes were implanted at the same coordinates where the virus was injected. The optrode was constructed by surrounding an optical fiber with tetrode wires and the tip of the optical fiber was about 200 μm above the tetrode tips. Blue light pulses (473 nm, 2-ms duration, 20 Hz) were delivered at the end of each recording session at high frequencies. Units were considered as light-responsive if they exhibited time-locked spiking with high reliability (> 90%), short first-spike latency (< 3 ms) and low jitter (< 2 ms) after light-pulse illumination. Only when the waveforms of laser-evoked and spontaneous spikes were highly similar (correlation coefficient > 0.9) were they considered to originate from a single neuron.

In vitro electrophysiological recordings

Acute brain slices preparation

After the acetic acid-treated mice were placed in the same or different context for 30 min on D2, then they were anesthetized with pentobarbital (20 mg/kg, i.p.). After intracardial perfusion with ~20 mL ice-cold NMDG ACSF contained the following (in mM): 93 NMDG, 20 HEPES buffer, 25 glucose, 2.5 KCl, 0.5 CaCl₂, 1.2 NaH₂PO₄, 30 NaHCO₃, 10 MgSO₄, 5 sodium ascorbate, 3 sodium pyruvate, 2 thiourea and 3 glutathione (pH: 7.3‒7.4, osmolarity: 300–305 mOsm/kg), coronal slices of the mPOA (300 μm) were sectioned in chilled (2‒4 °C) NMDG ACSF on a microtome (VT1200S, Leica, Germany) vibrating at 0.18 mm/s velocity. After that, the brain slices were initially incubated in NMDG ACSF (saturated with 95% O₂/5% CO₂ to provide stable pH and continuous oxygenation) for ~12 min at 33 °C which was followed by recovery for at least 1 hr at 25 °C in (HEPES) ACSF containing (in mM): 92 NaCl, 2.5 KCl, 2 CaCl₂, 1.2 NaH₂PO₄, 30 NaHCO₃, 5 Na-ascorbate, 3 Na-pyruvate, 25 glucose, 20 HEPES, 2 MgSO₄, 3 GSH and 2 thiourea (pH: 7.3‒7.4, osmolarity: 300‒310 mOsm/kg). The brain slices were transferred to a slice chamber (Warner Instruments, USA) for electrophysiological recording and were continuously perfused with standard ACSF that contained (in mM): 129 NaCl, 2.4 CaCl₂, 3 KCl, 1.3 MgSO₄, 20 NaHCO₃, 1.2 KH₂PO₄ and 10 glucoses (pH: 7.3–7.4, osmolarity: 300–305 mOsm/kg) with a flow rate of 2.5–3 mL/min. The temperature of the standard ACSF was maintained at 32 °C by an in-line solution heater (TC-344B, Warner Instruments, USA). During slice preparation and electrophysiology recording, all solutions were continuously bubbled with 95% O₂/5% CO₂.

Whole-cell patch-clamp recordings

Neurons were visualized with a 40× water-immersion objective on an infrared (IR)-differential interference contrast (DIC) microscope (BX51WI, Olympus, Japan) and an infrared camera was connected to the video monitor. Recording pipettes (3‒5 MΩ) were pulled from borosilicate glass capillaries (VitalSense Scientific Instruments Co., Ltd., Wuhan, China) with an outer diameter of 1.5 mm on a four-stage horizontal puller (P-1000, Sutter Instruments, USA). Whole-cell patch-clamp recordings were controlled by a multiClamp 700B amplifier combined with low-pass filtering at 2.8 kHz and digitized at 10 kHz. To record the activity of the Glu^mPOA neurons, tdTomato⁺ neurons were recorded in the mPOA of both male and female VgluT2-Ai14 mice. In order to test the effects of TP on neuronal excitability ex vivo, coronal slices containing the mPOA from C57 mice were prepared for whole-cell patch-clamp recordings. After 4 min of baseline recording for the acting potential, 1 μM TP (dissolved in DMSO) was added into the ACSF. The mPOA neurons from one side were recorded for at least 10 min in follow to observe the effects of TP, and other sides of mPOA neurons were also recorded separately with 1 μM saline (dissolved in DMSO) in the ACSF for control. Neurons with a series resistance of more than 30 MΩ or that changed by more than 20% during the recording were excluded.

Spontaneous inhibitory and excitatory postsynaptic currents (sIPSCs and sEPSCs)

The mPOA neurons were held at −70 mV using the voltage-clamp mode for recording sIPSCs or sEPSCs in the presence of the AMPA receptor blocker CNQX (10 μM) or the GABA_A receptor blocker BIC (10 μM). The recording pipettes (3–5 MΩ) were filled with a solution containing (in mM): 120 KCl, 30 NaCl, 10 HEPES, 1 MgCl₂, 0.5 CaCl₂, 2 Mg-ATP, 0.3 Na-GTP, and 5 EGTA for sIPSCs. The osmolarity of the solution was adjusted to 285-290 mOsm/kg, and the pH was adjusted to 7.2–7.3 with KOH (1 M). For sEPSCs, the pipettes filled with a solution containing (in mM): 130 potassium gluconate, 5 KCl, 2 MgCl₂, 2 Mg-ATP, 0.6 EGTA, 10 HEPES, and 0.3 Na-GTP (pH: 7.2-7.3, osmolality of 285-290 mOsm/kg). The signals were collected via a Multiclamp 700B amplifier and analyzed with Clampfit 10.7 software and Minianalysis Program 6.0 (Molecular Devices). Unless otherwise stated, all drugs used for electrophysiological recordings were purchased from Sigma-Aldrich (USA).

Light-evoked response

A laser (Shanghai Fiblaser Technology, China) through an optical fiber was used to deliver optical stimulation with positioned 0.2-mm above the surface of the target brain region. To verify the functionality of the AAV-DIO-ChR2-mCherry virus, mCherry-labelled neurons that expressed ChR2 in VgluT2-Cre mice were visualized and subjected to blue laser light (473 nm, 10 V, 20 ms) using 1-Hz, 5-Hz, and 10-Hz stimulation. The EPSCs were recorded with the membrane potential held at −70 mV. In some experiments, the functional characteristics of AAV-DIO-eNpHR3.0-EYFP virus were assessed by stimulation with a sustained yellow laser light. To confirm the function of Glu^mPOA → GABA^vlPAG pathway, AAV-CaMKII-ChR2-mCherry were injected into the mPOA and AAV-DIO-EGFP injected into the vlPAG of Vgat-Cre mice. Light-evoked postsynaptic currents of GABA^vlPAG neurons were elicited by 20-ms blue light stimulation of ChR2-expressing terminals from Glu^mPOA neurons that projecting to the vlPAG.

Statistics and reproducibility

OriginPro 2017 software (OriginLab Corporation, USA) and GraphPad Prism 8 (GraphPad Software, Inc., USA) were used for the statistical analyses and graphing. Offline analysis of the data obtained from electrophysiological recordings was conducted using Clampfit software version 10.7 (Axon Instruments, Inc., USA). For representative images and statistical data, each experiment has been repeated at least three mice with multiple brain slices in the manuscript with consistent results. We conducted statistical comparisons between two groups using paired or unpaired Student’s t-tests. One-way or Two-way analysis of variance (ANOVA) and Bonferroni post hoc analyses were used in analyses with multiple experimental groups. Data are shown as individual values or expressed as the means ± SEMs and significance levels are indicated as *P < 0.05, ** P < 0.01, *** P < 0.001 and not significant (n.s.). P values are not provided as exact values when they are less than 0.0001. All statistical tests, significance analyses, number of individual experiments (n) and other relevant information for data comparison are specified in Supplementary Table 1.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data necessary to understand and assess the conclusions of this study are available in the main text or the supplementary materials. There are no restrictions on data availability in the paper. Source data are provided with this paper.

References

Mogil, J. S. Qualitative sex differences in pain processing: emerging evidence of a biased literature. Nat. Rev. Neurosci. 21, 353–365 (2020).
Article PubMed CAS Google Scholar
Mogil, J. S. Sex differences in pain and pain inhibition: multiple explanations of a controversial phenomenon. Nat. Rev. Neurosci. 13, 859–866 (2012).
Article PubMed CAS Google Scholar
Navratilova, E., Fillingim, R. B. & Porreca, F. Sexual dimorphism in functional pain syndromes. Sci. Transl. Med 13, eabj7180 (2021).
Article PubMed CAS Google Scholar
Dawes, J. M. & Bennett, D. L. Addressing the gender pain gap. Neuron 109, 2641–2642 (2021).
Article PubMed CAS Google Scholar
Martin, L. J. et al. Male-specific conditioned pain hypersensitivity in mice and humans. Curr. Biol. 29, 192–201 e194 (2019).
Article PubMed CAS Google Scholar
Trask, S., Mogil, J. S., Helmstetter, F. J., Stucky, C. L. & Sadler, K. E. Contextual control of conditioned pain tolerance and endogenous analgesic systems. Elife 11, e75283 (2022).
Article PubMed PubMed Central CAS Google Scholar
Gulati, M., Dursun, E., Vincent, K. & Watt, F. E. The influence of sex hormones on musculoskeletal pain and osteoarthritis. Lancet Rheumatol. 5, e225–e238 (2023).
Article PubMed CAS Google Scholar
Blackburn-Munro, G. & Blackburn-Munro, R. Pain in the brain: are hormones to blame? Trends Endocrinol. Metab. 14, 20–27 (2003).
Article PubMed CAS Google Scholar
Fillingim, R. B. & Ness, T. J. Sex-related hormonal influences on pain and analgesic responses. Neurosci. Biobehav Rev. 24, 485–501 (2000).
Article PubMed CAS Google Scholar
Xie, Z. et al. Estrogen metabolites increase nociceptor hyperactivity in a mouse model of uterine pain. JCI Insight 7 e149107 (2022).
Vacca, V. et al. Higher pain perception and lack of recovery from neuropathic pain in females: a behavioural, immunohistochemical, and proteomic investigation on sex-related differences in mice. Pain 155, 388–402 (2014).
Article PubMed Google Scholar
Craft, R. M. Modulation of pain by estrogens. Pain 132, S3–S12 (2007).
Article PubMed CAS Google Scholar
Sorge, R. E. et al. Different immune cells mediate mechanical pain hypersensitivity in male and female mice. Nat. Neurosci. 18, 1081–1083 (2015).
Article PubMed PubMed Central CAS Google Scholar
Claiborne, J., Nag, S. & Mokha, S. S. Activation of opioid receptor like-1 receptor in the spinal cord produces sex-specific antinociception in the rat: estrogen attenuates antinociception in the female, whereas testosterone is required for the expression of antinociception in the male. J. Neurosci. 26, 13048–13053 (2006).
Article PubMed PubMed Central CAS Google Scholar
Sorge, R. E. et al. Spinal cord Toll-like receptor 4 mediates inflammatory and neuropathic hypersensitivity in male but not female mice. J. Neurosci. 31, 15450–15454 (2011).
Article PubMed PubMed Central CAS Google Scholar
White, H. D. & Robinson, T. D. A novel use for testosterone to treat central sensitization of chronic pain in fibromyalgia patients. Int. Immunopharmacol. 27, 244–248 (2015).
Article PubMed CAS Google Scholar
Basaria, S. et al. Effects of testosterone replacement in men with opioid-induced androgen deficiency: a randomized controlled trial. Pain 156, 280–288 (2015).
Article PubMed PubMed Central CAS Google Scholar
Lewis, S. S. et al. Select steroid hormone glucuronide metabolites can cause toll-like receptor 4 activation and enhanced pain. Brain Behav. Immun. 44, 128–136 (2015).
Article PubMed CAS Google Scholar
Vincent, K., Moore, J., Kennedy, S. & Tracey, I. Steroid hormones and pain-related brain activity and functional connectivity in healthy women. Lancet 383, S104 (2014).
Article Google Scholar
Wei, Y. C. et al. Medial preoptic area in mice is capable of mediating sexually dimorphic behaviors regardless of gender. Nat. Commun. 9, 279 (2018).
Article ADS PubMed PubMed Central Google Scholar
Yang, C. F. & Shah, N. M. Representing sex in the brain, one module at a time. Neuron 82, 261–278 (2014).
Article PubMed PubMed Central CAS Google Scholar
Zhang, G.-W. et al. Medial preoptic area antagonistically mediates stress-induced anxiety and parental behavior. Nat. Neurosci. 24, 516–528 (2021).
Article PubMed PubMed Central CAS Google Scholar
Goldman, A. L. et al. A reappraisal of testosterone’s binding in circulation: physiological and clinical implications. Endocr. Rev. 38, 302–324 (2017).
Article PubMed PubMed Central Google Scholar
Butler, R. K. & Finn, D. P. Stress-induced analgesia. Prog. Neurobiol. 88, 184–202 (2009).
Article PubMed CAS Google Scholar
Ford, G. K. & Finn, D. P. Clinical correlates of stress-induced analgesia: evidence from pharmacological studies. Pain 140, 3–7 (2008).
Article PubMed Google Scholar
LeBlancq, M. J., McKinney, T. L. & Dickson, C. T. ZIP it: neural silencing is an additional effect of the PKM-zeta inhibitor zeta-inhibitory peptide. J. Neurosci. 36, 6193–6198 (2016).
Article PubMed PubMed Central CAS Google Scholar
Sacktor, T. C. & Fenton, A. A. Appropriate application of ZIP for PKMζ inhibition, LTP reversal, and memory erasure. Hippocampus 22, 645–647 (2012).
Article PubMed CAS Google Scholar
Todd, A. J. Neuronal circuitry for pain processing in the dorsal horn. Nat. Rev. Neurosci. 11, 823–836 (2010).
Article PubMed PubMed Central CAS Google Scholar
Bliss, T. V., Collingridge, G. L., Kaang, B. K. & Zhuo, M. Synaptic plasticity in the anterior cingulate cortex in acute and chronic pain. Nat. Rev. Neurosci. 17, 485–496 (2016).
Article PubMed CAS Google Scholar
Zhu, X. et al. Distinct thalamocortical circuits underlie allodynia induced by tissue injury and by depression-like states. Nat. Neurosci. 24, 542–553 (2021).
Article PubMed CAS Google Scholar
Cao, P. et al. Green light induces antinociception via visual-somatosensory circuits. Cell Rep. 42, 112290 (2023).
Article PubMed CAS Google Scholar
Frew, A. K. & Drummond, P. D. Negative affect, pain and sex: the role of endogenous opioids. Pain 132, S77–S85 (2007).
Article PubMed CAS Google Scholar
Pisanu, C. et al. Sex differences in the response to opioids for pain relief: a systematic review and meta-analysis. Pharm. Res. 148, 104447 (2019).
Article CAS Google Scholar
Chia, Y. Y. et al. Gender and pain upon movement are associated with the requirements for postoperative patient-controlled iv analgesia: a prospective survey of 2,298 Chinese patients. Can. J. Anaesth. 49, 249–255 (2002).
Article PubMed Google Scholar
Calderone, K. L. The influence of gender on the frequency of pain and sedative medication administered to postoperative patients. Sex. Roles 23, 713–725 (1990).
Article Google Scholar
Gegenhuber, B., Wu, M. V., Bronstein, R. & Tollkuhn, J. Gene regulation by gonadal hormone receptors underlies brain sex differences. Nature 606, 153–159 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Brown, R. S. E. et al. Prolactin action in the medial preoptic area is necessary for postpartum maternal nursing behavior. Proc. Natl Acad. Sci. USA 114, 10779–10784 (2017).
Article ADS PubMed PubMed Central CAS Google Scholar
Tao, C. et al. The medial preoptic area mediates depressive-like behaviors induced by ovarian hormone withdrawal through distinct GABAergic projections. Nat. Neurosci. 26, 1529–1540 (2023).
Article PubMed PubMed Central CAS Google Scholar
Karigo, T. et al. Distinct hypothalamic control of same- and opposite-sex mounting behaviour in mice. Nature 589, 258–263 (2021).
Article ADS PubMed CAS Google Scholar
Yang, S. et al. An mPOA-ARC(AgRP) pathway modulates cold-evoked eating behavior. Cell Rep. 36, 109502 (2021).
Article PubMed CAS Google Scholar
Millan, M. J. Descending control of pain. Prog. Neurobiol. 66, 355–474 (2002).
Article PubMed CAS Google Scholar
François, A. et al. A brainstem-spinal cord inhibitory circuit for mechanical pain modulation by GABA and enkephalins. Neuron 93, 822–839.e826 (2017).
Article PubMed PubMed Central Google Scholar
Akaike, A., Shibata, T., Satoh, M. & Takagi, H. Analgesia induced by microinjection of morphine into, and electrical stimulation of, the nucleus reticularis paragigantocellularis of rat medulla oblongata. Neuropharmacology 17, 775–778 (1978).
Article PubMed CAS Google Scholar
Bobeck, E. N., Chen, Q., Morgan, M. M. & Ingram, S. L. Contribution of adenylyl cyclase modulation of pre- and postsynaptic GABA neurotransmission to morphine antinociception and tolerance. Neuropsychopharmacology 39, 2142–2152 (2014).
Article PubMed PubMed Central CAS Google Scholar
Fields, H. State-dependent opioid control of pain. Nat. Rev. Neurosci. 5, 565–575 (2004).
Article PubMed CAS Google Scholar
Tonsfeldt, K. J. et al. Sex Differences in GABAA Signaling in the Periaqueductal Gray Induced by Persistent Inflammation. J. Neurosci. 36, 1669–1681 (2016).
Article PubMed PubMed Central CAS Google Scholar
Xie, L. et al. Divergent modulation of pain and anxiety by GABAergic neurons in the ventrolateral periaqueductal gray and dorsal raphe. Neuropsychopharmacology 48, 1509–1519 (2023).
Article PubMed CAS Google Scholar
Hohmann, A. G. et al. An endocannabinoid mechanism for stress-induced analgesia. Nature 435, 1108–1112 (2005).
Article ADS PubMed CAS Google Scholar
McEwen, B. S. & Kalia, M. The role of corticosteroids and stress in chronic pain conditions. Metabolism 59, S9–S15 (2010).
Article PubMed CAS Google Scholar
Toledo-Rodriguez, M. & Sandi, C. Stress before puberty exerts a sex- and age-related impact on auditory and contextual fear conditioning in the rat. Neural Plast. 2007, 71203 (2007).
Article PubMed PubMed Central Google Scholar
Sanacora, G., Yan, Z. & Popoli, M. The stressed synapse 2.0: pathophysiological mechanisms in stress-related neuropsychiatric disorders. Nat. Rev. Neurosci. 23, 86–103 (2022).
Article PubMed CAS Google Scholar
Pang, H. et al. Low back pain and osteoarthritis pain: a perspective of estrogen. Bone Res. 11, 42 (2023).
Article PubMed PubMed Central CAS Google Scholar
Kaur, S. et al. Estrogen modulation of the pronociceptive effects of serotonin on female rat trigeminal sensory neurons is timing dependent and dosage dependent and requires estrogen receptor alpha. PAIN 163, e899–e916 (2022).
Article PubMed PubMed Central CAS Google Scholar
Hays, J. et al. Effects of estrogen plus progestin on health-related quality of life. N. Engl. J. Med. 348, 1839–1854 (2003).
Article PubMed CAS Google Scholar
Berkley, K. J., Rapkin, A. J. & Papka, R. E. The pains of endometriosis. Science 308, 1587–1589 (2005).
Article ADS PubMed CAS Google Scholar
Silva, R. F. F. et al. The anabolic androgenic steroid testosterone propionate decreases recognition memory in adult male rats. Curr. Psychopharmacol. 2, 247–253 (2013).
Article CAS Google Scholar
Clark, A. S. & Barber, D. M. Anabolic-androgenic steroids and aggression in castrated male rats. Physiol. Behav. 56, 1107–1113 (1994).
Article PubMed CAS Google Scholar
Parrilla-Carrero, J. et al. The anabolic steroids testosterone propionate and nandrolone, but not 17alpha-methyltestosterone, induce conditioned place preference in adult mice. Drug Alcohol Depend. 100, 122–127 (2009).
Article PubMed CAS Google Scholar
King, B. E., Packard, M. G. & Alexander, G. M. Affective properties of intra-medial preoptic area injections of testosterone in male rats. Neurosci. Lett. 269, 149–152 (1999).
Article PubMed CAS Google Scholar
Dong, Q., Rintala, H. & Handelsman, D. J. Androgen receptor function during undernutrition. J. Neuroendocrinol. 6, 397–402 (1994).
Article PubMed CAS Google Scholar

Download references

Acknowledgements

This work was supported by the National Key Research and Development Program of China (STI2030-Major Projects 2021ZD0203100 to Z. Z.), the Plans for Major Provincial Science & Technology Projects (202303a07020002 to Z. Z.), the National Natural Science Foundation of China (32025017 to Z. Z., 32121002 to Z. Z., 32400824 to P. C.), the Anhui Provincial Natural Science Foundation (2308085QH264 to P. C.), CAS Project for Young Scientists in Basic Research (YSBR-013 to Z. Z.), Research Funds of Center for Advanced Interdisciplinary Science and Biomedicine of IHM (QYZD20220007 to Z. Z.), the Fundamental Research Funds for the Central Universities (WK9100000068 to P. C.), the USTC Research Funds of the Double First-Class Initiative (YD9110002015 to J. L.).

Author information

These authors contributed equally: Mingjun Zhang, Ziyun Ni, Jun Ma, An Liu.

Authors and Affiliations

Department of Anesthesiology, The First Affiliated Hospital of USTC, Center for Advanced Interdisciplinary Science and Biomedicine of IHM, Hefei National Laboratory for Physical Sciences at the Microscale, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei, 230026, China
Mingjun Zhang, Ziyun Ni, Jun Ma, Ying Liu, Qianqian Lou, Wan-Ying Dong, Zhi Zhang, Juan Li & Peng Cao
Department of Physiology, School of Basic Medical Sciences, Anhui Medical University, Hefei, 230022, China
An Liu & Zhi Zhang
Department of Biophysics and Neurobiology, CAS Key laboratory of Brain Function and Disease, University of Science and Technology of China, Hefei, 230026, China
Zhi Zhang & Peng Cao

Authors

Mingjun Zhang
View author publications
Search author on:PubMed Google Scholar
Ziyun Ni
View author publications
Search author on:PubMed Google Scholar
Jun Ma
View author publications
Search author on:PubMed Google Scholar
An Liu
View author publications
Search author on:PubMed Google Scholar
Ying Liu
View author publications
Search author on:PubMed Google Scholar
Qianqian Lou
View author publications
Search author on:PubMed Google Scholar
Wan-Ying Dong
View author publications
Search author on:PubMed Google Scholar
Zhi Zhang
View author publications
Search author on:PubMed Google Scholar
Juan Li
View author publications
Search author on:PubMed Google Scholar
Peng Cao
View author publications
Search author on:PubMed Google Scholar

Contributions

Conceptualization: M. Z., P. C., and Z. Z.; Methodology: M. Z., Z. N., Y. L., and W. D.; Investigation: M. Z., Z. N., J. M., and A. L.; Visualization: M. Z., Q. L., and A. L.; Funding acquisition: J. L., P. C., and Z. Z.; Project administration: Z. Z.; Supervision: Z. Z. and P. C.; Writing—original draft: M. Z. and P. C.; Writing—review & editing: J. L., P. C., and Z. Z.

Corresponding authors

Correspondence to Zhi Zhang, Juan Li or Peng Cao.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Zhang, M., Ni, Z., Ma, J. et al. A neural circuit for sex-dependent conditioned pain hypersensitivity in mice. Nat Commun 16, 3639 (2025). https://doi.org/10.1038/s41467-025-58851-x

Download citation

Received: 21 November 2024
Accepted: 01 April 2025
Published: 17 April 2025
Version of record: 17 April 2025
DOI: https://doi.org/10.1038/s41467-025-58851-x

This article is cited by

Efficacy and safety of intrathecal human umbilical cord-derived mesenchymal stromal cell therapy for bone cancer pain: a preclinical study
- Jing Zhang
- Yunqi Li
- Qingping Wen
Stem Cell Research & Therapy (2025)
Effect and mechanism of father’s companionship on defensive attack behavior of adult male offspring mice
- Yating Li
- Yiran Liu
- Haishui Shi
Communications Biology (2025)