Role of astrocytic mu-opioid receptors of the ventrolateral periaqueductal gray in modulating anxiety-like responses

doi:10.21203/rs.3.rs-6262877/v1

Download PDF

Research Article

Role of astrocytic mu-opioid receptors of the ventrolateral periaqueductal gray in modulating anxiety-like responses

https://doi.org/10.21203/rs.3.rs-6262877/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 23 Jul, 2025

Read the published version in Behavioral and Brain Functions →

You are reading this latest preprint version

Background

Mu-opioid receptors (MORs) are critical regulators mediating the modulation of several behavioral reactions, including analgesia, addiction, and sedation. Recent studies have reported that MORs are closely associated with mood disorders or anxiety behaviors; however, the underlying neural mechanisms remain unclear. The periaqueductal gray (PAG), a key brain area, participates in the modulation of aversive emotional behaviors. MORs show a high expression in the ventrolateral PAG (vlPAG) region. This study explored the preliminary role of MORs expressed in the vlPAG in modulating emotional behaviors.

Results

Bilateral administration of DAMGO, an MOR-specific agonist, into the vlPAG of male mice elicited anxiety-like behaviors in elevated plus maze tests. This phenotype was reversed by conditional knockdown of astrocytic MORs. In contrast, glutamatergic or GABAergic MORs were not involved in vlPAG MOR-dependent anxiety-like behaviors. By using in vitro calcium imaging of vlPAG astrocytes and chemical genetic technologies, we found that vlPAG astrocytic MORs can promote astrocytic calcium signaling, which can efficiently induce anxiety-like behaviors. Accordingly, the interference of astrocytic calcium signaling by viral infection reversed vlPAG-dependent anxiety-like behaviors.

Conclusion

Our findings demonstrated that vlPAG astrocytic, but not glutamatergic or GABAergic, MORs are involved in modulating emotional reactions, and these effects are accomplished by MOR-elicited astrocytic calcium signaling mechanisms. The present study provides a theoretical basis for treating emotional dysfunctions during MOR-targeted management.

Anxiety

mu-opioid receptors

periaqueductal gray

astrocytes

calcium signaling

Mu-opioid receptor (MOR), a member of the endogenous opioid system, is a major regulator of pain sensation and reaction in the central nervous system (CNS) [1, 2]. According to several studies, a strong correlation exists between MOR activity levels and anxiety behavior [3–5]. A high prevalence of mood disorders and anxiety behavior is observed in individuals habitually using MOR-targeted analgesics [4, 5]. Thus, the role of MORs in influencing anxiety or anxiety-like behaviors requires greater attention.

MORs are widely expressed in several brain areas that control emotional behaviors, such as the amygdala, hippocampus, prefrontal cortex, bed nucleus of the stria terminalis, and periaqueductal gray (PAG) [6, 7]. Among these areas, the PAG is a brainstem region with predominant involvement in modulating aversive emotional or defensive behaviors [8–10]. It receives afferent connections from the prefrontal cortex, central amygdala, hypothalamus, and habenula nucleus; it integrates neurotransmission from these brain areas and exports an emotional primary process [8, 10]. The PAG contains four longitudinal columns: the dorsomedial PAG (dmPAG), dorsolateral PAG (dlPAG), lateral PAG (lPAG), and ventrolateral PAG (vlPAG) [8–10]. The vlPAG has a close association with negative emotional responses [11, 12]. Altered excitability of vlPAG neurons has been observed in emotional-related disorders, and chemogenetic activation of vlPAG glutamatergic neurons induces anxiety-like behavior in the light/dark test and open field test in rodents [11]. These findings indicate the functional role of the vlPAG in modulating negative emotional behavior. Both morphological and biochemical investigations have reported the dense distribution of MORs in the vlPAG [2, 7]. Additionally, pharmacological or genetic intervention targeted toward the functions of vlPAG MORs can fluctuate the neural excitability of the PAG [7, 13]. These findings suggest a potential role of PAG MORs in modulating emotional behavior. Moreover, by using combined in situ hybridization and immunofluorescence technology, several cell types, including glutamatergic (MOR_Glut), gamma-aminobutyric acid (GABA) ergic (MOR_GABA), and astrocytic (MOR_Astro) neurocytes, were found to express MORs in the vlPAG [7, 14]; this suggests that the multiple interactions between MORs on these different neuronal types might be involved in emotional behavior control.

Here, we performed bilateral micro-administration of the MOR-specific agonist DAMGO in male mice and observed how local activation of vlPAG MORs affected anxiety-like behavior through the elevated plus maze (EPM) test. By developing three male murine models in which glutamatergic, GABAergic, or astrocytic MORs were conditionally knocked down, we detected the role of divergent neuronal types of vlPAG MORs on the modulation of anxiety-like behaviors. Furthermore, by using in vitro calcium imaging and chemogenetic and virus-interfering technologies, we elucidated the preliminary mechanisms of these effects. DAMGO induced significant MOR-dependent anxiety-like effects in the vlPAG. Interestingly, these effects were induced by MOR_Astro but not by MOR_Glut or MOR_GABA. Moreover, the induction and enhancement of calcium signaling could be critical mechanisms for inducing vlPAG MOR_Astro-dependent anxiety-like behavior. Our study highlights a new MOR_Astro elicited anxiety-like behavior paradigm in the PAG. These observations offer a theoretical premise for treating emotional dysfunctions during MOR-targeted management.

Animals

The study involved 78 C57BL/6J male mice, 42 MOR mutant male mice, and 42 littermate male controls (8–12 weeks old). All mice were supplied by the Experimental Animal Center of Shaanxi Normal University, MOE Key Laboratory of Modern Teaching Technology. The mice were housed in individually ventilated cages in groups (5–6 animals per group) and maintained under the following conditions: temperature: 21°C ± 2°C; relative humidity: 50% ± 5%; sufficient food and water; and 12-h light/dark cycle. Before the behavioral experiments, the animals were acclimated to the environment and apparatus. The use count of mice in each experiment is detailed in the Results section.

The mutant mouse lines were generated as reported previously [7, 15]. Briefly, mice specifically lacking MORs in MOR_Astro, MOR_GABA, or MOR_Glut were generated by crossing Oprm1-floxed mice to GFAP-CreERT2, GAD2-iCreERT2, or vGlut2-iCreERT2 mice, respectively. The adult Oprm1^loxP/loxP:GFAP-CreERT2 (MOR_Astro−/−), Oprm1^loxP/loxP:Gad2-iCreERT2 (MOR_GABA−/−), or Oprm1^loxP/loxP:vGlut2-iCreERT2 (MOR_Glut−/−) mice were intraperitoneally administered for 7 consecutive days with tamoxifen (2 mg/day; Sigma-Aldrich) for inducing MOR knockdown and subsequently utilized for the experiments at 2 weeks after the final injection. The littermates of these three types of mice (Oprm1-flox+/+:CreERT2−/−) receiving identical tamoxifen treatment were considered controls (MOR_Astro+/+, MOR_GABA+/+, or MOR_Glut+/+).

Surgery and intra-vlPAG injections

After inducing anesthesia with 4% isoflurane inhalant, mice were placed in a brain stereotaxic apparatus (RWD, China) for the surgical insertion of guide cannulas. The nasopharynx of each mouse was continuously treated with 1% isoflurane inhalant for maintaining anesthesia during surgery. After the skull was exposed, the guide cannulas (double-barreled, 0.8 mm apart; length: 4 mm, internal diameter: 0.34 mm, external diameter: 0.48 mm; RWD) were inserted into the bilateral of the vlPAG (site: anteroposterior: −4.80 mm, mediolateral: ± 0.40 mm, dorsoventral: −2.85 mm) according to the mouse brain atlas. The upper-half parts of the guide cannulas were secured to the skull through the cannulas and pre-covered with protective caps before intravenous PEG administration to prevent clogging. A diclofenac sodium gel was smeared near the wounds to achieve postoperative pain relief. After surgery, the mice were transferred to a newly ventilated cage with a clean and pathogen-free environment. All animals underwent recovery for at least 7 days before intra-vlPAG injection.

The intra-vlPAG injections were processed as described previously. Briefly, the drugs for injection were aspirated into 0.5 µL needle-tipped micro-syringes (the process was regulated by a micro-infusion pump; RWD). The micro-syringes were then connected with the injecting cannulas (double-barreled, 0.8 mm apart; length: 4 mm, internal diameter: 0.14 mm, external diameter: 0.30 mm; RWD) by using plastic hoses. The injecting cannulas were then inserted into guide cannulas. Subsequently, the drugs were intra-vlPAG injected through micro-infusion pump control (0.1 µL each side, 0.04 µL/min, duration: 3 min). The following drugs were used in the intra-vlPAG injections: DAMGO (100 µM, dissolved in normal saline, concentration based on a previous study [16]; Tocris, UK) and CTAP (1 mM, dissolved in normal saline; concentration based on a previous study [16]; Tocris). To minimize the effects of stress elicited by the injection procedures during behavioral tests, mice were subjected to daily simulated intracerebral injections for 3 consecutive days before behavioral tests. The injection sites of each mouse were morphologically examined after the behavioral test. Data of mice with injection sites different from vlPAG were rejected.

EPM test

The EPM testing device comprised two mutually orthogonal open arms and two closed arms (length × width: 28 × 5.8 cm) intersected at a central square (length × width: 5.8 × 5.8 cm). The two close arms were placed opposite to each other and were surrounded by 15.5-cm-high walls. The height of the maze was 55 cm above the ground. The image-capturing camera was placed directly above the central square of the EPM. During the test, the mouse was placed in the central square, and its head faced one of the open arms. Each mouse was allowed 5 min of free exploration. The total distance traveled, percentage of time spent in the open arms, and percentage of open arm entries were determined using Mouse EthoVision XT (Noldus, Holland). In each mouse, the percentage of open arm entries was estimated as follows: number of open arm entries/number of total arm entries × 100%; the percentage of time spent in the open arms was estimated as follows: time spent in open arms/time spent in all arms × 100%. After each test, the testing device was cleaned and disinfected with 75% ethanol.

Recombinant adeno-associated virus (rAAV) injection

The surgery and rAAV injection were performed as described previously. Briefly, after inducing anesthesia with 4% isoflurane inhalant, the mice were placed in the brain stereotaxic apparatus. After exposing the skull, the corresponding rAAVs were injected into the vlPAG at the rate of 30 nL/min. Following this injection, the head skin of each mouse was carefully sutured. The postoperative treatment was identical to that of embedding of the guide cannula. The following rAAVs were used: rAAV2/5-GfaABC1D-cyto-GCaMP6f-SV40 pA (100 nL/injection, 1.20 × 10¹³ vg/mL; Brain VTA, China) was used for fluorescent visualization of calcium activity of astrocytes in the vlPAG; rAAV2/5-GfaABC1D-hM3D(Gq)-mCherry-SV40 pA (170 nL/injection, 5.93 × 10¹² vg/mL; Brain VTA) was used to homogenetically activate the calcium signaling of astrocytes in the vlPAG; rAAV2/5-GfaABC1D-hPMCA2w/b-mCherry-SV40 pA (170 nL/injection, 5.27 × 10¹² vg/mL; Brain VTA) was used to specifically intercept calcium signaling of astrocytes in the vlPAG; and rAAV2/5-GfaABC1D-mCherry-SV40 pA (170 nL/injection, 5.27 × 10¹² vg/mL; Brain VTA) was used as a control for the above rAAVs. All animals underwent recovery for at least 21 days for ensuring complete expression of rAAVs.

In vitro calcium imaging of vlPAG astrocytes

The calcium activity of vlPAG astrocytes was monitored and analyzed by in vitro calcium imaging as described previously. Under 4% isoflurane-induced anesthesia, mice injected with rAAV2/5-GfaABC1D-cyto-GCaMP6f-SV40 pA were rapidly decapitated, and their brains were removed and rapidly transferred into oxygenated modified artificial cerebrospinal fluid (ACSF, including [in mM]: NaCl, 125; KCl, 2.5; NaH₂PO₄, 1.25; MgCl₂, 2; CaCl₂, 2; glucose, 25; and NaHCO₃, 25; pH 7.4). By using a vibrating slicer (1000 plus; Vibratome Company, St. Louis, MO, USA), 300-µm-thick coronal slices containing vlPAG were cut. The vlPAG astrocytes were imaged using continuous fluorescence excitation with a 488-nm light source (Leica, Germany). The imaging sessions were conducted at the rate of 1 frame/s. Brain slice processing and astrocyte imaging were performed under oxygenated ACSF incubation. The bath solution for imaging contained the following materials individually or mixed: DAMGO (1 µM), CTAP (10 µM), and/or tetrodotoxin (TTX, 1 µM; 554412, Sigma-Aldrich, USA). TTX and CTAP were added to the bath 10 min before the start of recording. Imaging data were acquired and analyzed using ImageJ software.

Morphological assessment of mutant mouse lines

To confirm the knockdown efficiency of MOR_Astro−/−, MOR_GABA−/−, and MOR_Glut−/− mice, fluorescence in situ hybridization (FISH) with RNAscope and immunofluorescence were used as described previously [7, 15]. Briefly, mice were deeply anesthetized with 4% isoflurane inhalant and subjected to cardiovascular perfusion with 0.9% saline for 5 min. The brains were rapidly removed and embedded in an optimal cutting temperature (OCT) compound (SAKURA Tissue-Tek, Japan) at − 22°C. A CM1950 freezing microtome (Leica, Germany) was utilized for cutting fresh frozen sections (16 µm) containing the vlPAG region (coronal plane). After fixation with 4% paraformaldehyde (PFA) for 30 min at 4°C, the sections were dehydrated using three grades of ethanol (50%, 75%, and 100%) for 5 min each at 25°C.

For fluorescence staining of MOR_Glut−/−, MOR_GABA−/− mice and their littermates, FISH alone was used. The sections were pretreated with hydrogen dioxide and protease IV for 10 and 15 min, respectively. The sections were incubated with probes for vglut2 (416631-C1, ACD, USA) conjugated to Atto 520, Oprm1 (544731-C2, ACD) conjugated to Atto 570, and GAT (424548-C3, ACD) conjugated to Atto 650. In situ hybridization was performed in a HybEZTM oven (ACD) by using an RNAscope Multiplex Fluorescent Reagent Kit (ACD) in accordance with the manufacturer’s protocol. Finally, the sections were mounted in a DAPI-containing anti-fade mounting medium. Both FISH with RNAscope and immunofluorescence were conducted for fluorescence staining of MOR_Astro−/− mice and littermates. Fluorescence labeling of Oprm1 was conducted following the protocol for FISH. The fluorescence labeling of astrocytes was performed by sequentially incubating the sections with rabbit anti-GFAP primary antibodies (1:500, 16825-1-AP, Proteintech, China) at 4°C (12 h) and dylight488-conjugated goat anti-rabbit antibodies (1:500, A23240, Abbkine, China) at room temperature (2 h). Subsequently, the sections were mounted with a DAPI-containing anti-fade mounting medium when both Oprm1 and astrocytes were labeled. A fluorescence microscope (Zeiss, Germany) was used to acquire confocal images of all sections, and cells showing positive labeling were enumerated.

Morphological examination for confirmation of the specific expression of rAAVs

To confirm the specific expression of rAAVs coupled with the GfaABC1D promoter in vlPAG astrocytes, immunofluorescence was performed as reported previously [17]. Briefly, mice injected with rAAVs were anesthetized by 4% isoflurane inhalant, before transcardial perfusion with 0.9% sodium chloride solution (Kelun, China) and 4% PFA fixative (BL539A, Biosharp, China). The brains were removed, post-fixed with 4% PFA fixative for 24 h, and then immersed in 30% sucrose phosphate-buffered saline (PBS) solution at 4°C (48 h). Subsequently, the brains were embedded in an OCT compound at -22°C. Fresh frozen sections (16 µm) containing the vlPAG region (coronal plane) were cut. For immunofluorescence assay, the sections were incubated for 60 min with PBS supplemented with 10% non-immune donkey serum (T8200, SolarBio, China) and 0.5% Triton X-100 (BL939A, Biosharp, China). The sections were sequentially incubated with rabbit anti-GFAP antibodies (1:300, 16825-1-AP, Proteintech) at 4°C (12 h) and dylight680-conjugated goat anti-rabbit antibodies (1:500, A23720, Abbkine) at room temperature (2 h). After labeling, the sections were confocally imaged using a fluorescence microscope to acquire confocal images of the sections, and cells showing positive labeling were enumerated.

Statistical analysis

Data are expressed in their original form or as mean ± SEM. GraphPad Prism 9.0 was utilized for data analysis. Comparison of two groups was achieved through unpaired or paired Student’s t test, and comparison of multiple groups was performed with one-way analysis of variance (ANOVA) and Sidak’s multiple comparison test. Statistical significance was considered at p < 0.05.

vlPAG MOR activation triggered anxiety-like behavior

MORs play a role in modulating emotional responses; however, the specific brain mechanisms underlying this effect remain unclear. To determine how MORs expressed on the vlPAG modulate anxiety-like behavior, we conducted bilateral insertion of guide cannulas into the vlPAG of mice (Figs. 1A, B) to specifically activate vlPAG MORs through pharmacological administration. The cannula position was confirmed morphologically (Fig. 1B). Next, we determined the effect of intravenous DAMGO, saline, or DAMGO + CTAP on anxiety-like behaviors. We found that the total distance traveled by mice in the EPM tests was comparable between the three mice groups, suggesting that locomotor activity was not affected by DAMGO (Figs. 1C, D). However, intra-vlPAG DAMGO induced a significant anxiogenic performance, as shown by the lower percentage of open arm entries and time spent in the open arms in control mice than in those with intra-vlPAG saline administration (Figs. 1C, E, F). These effects were reversed by CTAP (Figs. 1C, E, F), thus indicating the functional role of vlPAG MORs in modulating anxiety responses.

MOR_Astro, but not MOR_GABA or MOR_Glut, is involved in vlPAG-MOR-induced anxiety-like behavior

In the vlPAG, MORs exhibit a high expression level in different cell types such as astrocytes, GABAergic neurons, and glutaminergic neurons [7, 14]. To assess the distinct effects of vlPAG MOR_Astro, MOR_GABA, and MOR_Glut on the vlPAG-MOR elicited anxiety-like behavior, three mouse lines specifically lacking MOR_Astro−/−, MOR_GABA−/−, and MOR_Glut−/−, respectively, and their controls (MOR_Astro+/+, MOR_GABA+/+, or MOR_Glut+/+, respectively) were generated according to previously reported methods. To further confirm the validity of MOR deletions, the absence of astrocytic MORs of MOR_Astro−/− mice (Figs. 2A, B), GABAergic MORs of MOR_GABA−/− mice (Figs. 3A, B), and glutamatergic MORs of MOR_Glut−/− mice (Figs. 4A, B) were morphologically detected by combined FISH with RNAscope and immunofluorescence technology. We inserted guide cannulas into the bilateral vlPAG of these mice and tested anxiety-like behavior through the EPM test after intra-vlPAG DAMGO or saline administration.

During assessment of the effects of vlPAG MOR_Astro on the vlPAG-MOR-elicited anxiety-like behavior, EPM tests showed that the total distance traveled by mice was slightly different among the four mice groups, suggesting that genetic absence of MOR_Astro did not affect locomotor activity (Fig. 2C). MOR_Astro−/− and MOR_Astro+/+ mice with intra-vlPAG saline administration (Figs. 2D, E) showed no significant differences in the percentage of open arm entries and time spent in the open arms, indicating that the absence of MOR_Astro did not alter the basal emotional responses of mice. However, the anxiogenic effects induced by vlPAG MORs were significantly reversed by MOR_Astro−/−, as shown by the reduced percentage of open arm entries and time spent in the open arms of MOR_Astro+/+ mice with intra-vlPAG DAMGO administration compared to that of MOR_Astro+/+ mice with intra-vlPAG saline administration; in contrast, no significant differences were noted in the percentage of open arm entries and time spent in the open arms between MOR_Astro−/− mice with intra-vlPAG DAMGO administration and MOR_Astro−/− mice with intra-vlPAG saline administration (Figs. 2D, E). A remarkable difference was found in the percentage of open arm entries and time spent in the open arms between MOR_Astro+/+ mice administered with intra-vlPAG DAMGO administration and MOR_Astro−/− mice with intra-vlPAG DAMGO administration (Figs. 2D, E). Thus, MORs_Astro have a crucial role in vlPAG-MOR-induced anxiety-like behavior.

We then assessed the effects of vlPAG MOR_GABA on vlPAG-MOR-elicited anxiety-like behavior and found that the genetic absence of MOR_GABA did not affect locomotor activity; this was confirmed by the observation that the total distance traveled by mice was slightly different among the four mice groups (Fig. 3C). The percentage of open arm entries and time spent in the open arms displayed no significant differences between MOR_GABA−/− and MOR_GABA+/+ mice with intra-vlPAG saline administration (Figs. 3D, E), indicating that the absence of MOR_GABA did not alter the basal emotional responses of mice. Both MOR_GABA−/− and MOR_GABA+/+ mice were anxiogenic after intra-vlPAG DAMGO administration, and no significant difference was noted between MOR_GABA+/+ mice with intra-vlPAG DAMGO administration and MOR_GABA−/− mice with intra-vlPAG DAMGO administration (Figs. 3D, E). Thus, MOR_GABA hardly participates in vlPAG-MOR-induced anxiety-like behavior.

Similar to MOR_GABA, the assessment of how vlPAG MOR_GABA affects vlPAG-MOR-elicited anxiety-like behavior revealed no influence of the genetic absence of MOR_Glut on locomotor activity (Fig. 4C). No significant variations were noted in the percentage of open arm entries and time spent in the open arms between MOR_Glut−/− and MOR_Glut+/+ mice following intra-vlPAG saline administration (Figs. 4D, E), thus indicating that the absence of MOR_Glut did not alter the basal emotional responses of mice. Both MOR_Glut−/− and MOR_Glut+/+ mice were anxiogenic after intra-vlPAG DAMGO administration, and there was no significant difference between MOR_Glut+/+ mice with intra-vlPAG DAMGO administration and MOR_Glu−/− mice with intra-vlPAG DAMGO administration (Figs. 4D, E). Thus, MOR_Glut barely participates in vlPAG-MOR-induced anxiety-like behavior.

Taken together, our results indicate that MOR_Astro, but not MOR_GABA or MOR_Glut, is involved in vlPAG-MOR-induced anxiety-like behavior.

vlPAG MOR_Astro induces anxiety-like behavior through astrocytic calcium signaling

Next, to investigate the cellular mechanisms underlying MOR_Astro-dependent anxiety-like behavior, we focused on the role of vlPAG MOR_Astro activation and its effects on intracellular calcium signaling in astrocytes. The activation of MOR_Astro can trigger calcium signaling in the CNS [18]. As described previously [19, 20], GCamp6f was expressed in PAG astrocytes by using a viral vector with the GfaABC1D promoter for calcium imaging (Fig. 5A). Immunohistochemical assay confirmed the predominant expression of GCamp6f in astrocytes (Figs. 5B, C). The treatment of vlPAG slices with the specific MOR agonist DAMGO (1 µM) significantly increased calcium levels in vlPAG astrocytes (Figs. 5D, F); this was not observed in the control group perfused with DAMGO-free ACSF (Fig. 5E). The increase in calcium levels was inhibited by CTAP (10 µM) (Fig. 5G) but not by TTX (1 µM) (Fig. 5H), suggesting that this increase was specifically mediated by MOR_Astro activation.

To further investigate the involvement of astrocytic calcium signaling in the vlPAG in modulating anxiogenic performance, we examined whether direct activation of PAG astrocytes alone is sufficient to produce anxiety-like behavior. Designer receptors exclusively activated by designer drugs (DREADDs) coupled with the GfaABC1D promotor virus were used, which efficiently activated the calcium signals of astrocytes, as described previously [21, 22]. We injected rAAV5-GfaABC1D-hM3Dq-mCherry-WPRE-pA into the vlPAG targeting astrocytes (Astro_hM3Dq), with rAAV5-GfaABC1D-mCherry-WPRE-pA (Astro_control) as the control (Fig. 6A). Immunofluorescence results showed that the mCherry report proteins and GFAP were mostly co-labeled (Figs. 6B, C), indicating that hM3Dq rAAVs were specifically expressed in PAG astrocytes. The intraperitoneal administration of 1 mg/kg CNO remarkably caused anxiety-like behavior in Astro_hM3Dq mice; this was confirmed by the lower percentage of open arm entries and time spent in the open arms as compared to those of Astro_control mice with CNO administration or Astro_hM3Dq mice with saline administration (Figs. 6E, F). The total distance traveled by mice was comparable between these three mice groups, suggesting no influence of Astro_hM3Dq on locomotor activity (Fig. 6D). Thus, the direct activation of vlPAG astrocytic calcium signaling elicits anxiety-like behavior.

Next, we examined whether MOR_Astro-triggered calcium signaling is the cellular mechanism underlying the vlPAG MOR_Astro-dependent anxiety-like behavior. For this purpose, we bilaterally delivered rAAV5-GfaABC1D-hPMCA2w/b-mCherry-WPRE-SV40 into the PAG (Astro_hPMCA2w/b) (Fig. 7A), which specifically extruded cytoplasmic Ca²⁺ and reduced Ca²⁺ oscillations in PAG astrocytes, as described previously [23–25]. The rAAV5-GfaABC1D-mCherry-WPRE-SV40 virus was also injected into the PAG to generate control mice (Astro_control). The results of immunofluorescence assay showed that the mCherry report proteins and GFAP were mostly co-labeled (Figs. 7B, C), indicating that hPMCA2w/b rAAVs were specifically expressed in PAG astrocytes. We inserted guide cannulas into the bilateral vlPAG of these mice and tested their anxiety-like behavior through EPM tests following intra-vlPAG DAMGO or saline administration. Astro_hPMCA2w/b and Astro_control mice with intra-vlPAG saline administration showed no apparent differences in the percentage of open arm entries and time spent in the open arms (Figs. 7E, F); this finding indicated that Astro_hPMCA2w/b did not alter the basal emotional responses of mice. However, the anxiogenic performance elicited by vlPAG MORs were significantly reversed by Astro_hPMCA2w/b, as shown by the decreased percentage of open arm entries and time spent in the open arms of Astro_control mice with intra-vlPAG DAMGO administration as compared to those of Astro_control mice with intra-vlPAG saline administration. In contrast, Astro_hPMCA2w/b mice with intra-vlPAG DAMGO administration and Astro_hPMCA2w/b mice with intra-vlPAG saline administration showed no significant differences in the two abovementioned parameters (Figs. 7E, F). The percentage of open arm entries as well as the time spent in the open arms displayed significant differences between Astro_control mice with intra-vlPAG DAMGO administration and Astro_hPMCA2w/b mice with intra-vlPAG DAMGO administration (Figs. 7E, F). These four mice groups showed comparable total distance traveled, suggesting that the locomotor activity was not influenced by Astro_hPMCA2w/b (Fig. 6D). Thus, astrocytic calcium signaling is the cellular mechanism underlying vlPAG MOR_Astro-dependent anxiety-like behavior.

In summary, the obtained results indicate that vlPAG MOR_Astro elicits anxiety-like behavior through astrocytic calcium signaling.

Here, we examined the role of PAG MORs in modulating anxiety-like behavior. We observed that the local activation of PAG MORs through microinjection of DAMGO in the PAG elicited an apparent MOR-dependent anxiety-like behavior in the EPM tests. Surprisingly, neither MOR_Glut nor MOR_GABA was involved in modulating this form of MOR-dependent anxiety-like behavior. In contrast, in the EPM test, conditional knockdown of MOR_Astro reversed the DAMGO-induced reduction in the percentage of open arm entries and time spent in the open arms. By using calcium imaging and chemogenetic technologies, we further demonstrated the critical role of MOR_Astro in inducing astrocytic calcium signaling during the modulation of MOR_Astro-dependent anxiety-like behavior. Our study demonstrates a novel MOR_Astro-dependent emotional response in the PAG, which reveals a possible solution for treating emotional dysfunctions during MOR-targeted management in clinical settings.

The functional role of MORs in analgesia has been widely studied, with corresponding theories implemented in the area of clinical pain management [1, 2]. However, several clinical studies have cautioned about the increased risks of emotional dysfunction among patients who use opioid drugs during pain management [3–5]. In addition to pain modulation, MORs are extensively involved in emotional responses [26–28]. In clinical settings, patients with major depressive disorder display lower MOR availability [3]. Indeed, according to previous studies, blocking the effect of MORs by oral administration of naltrexone increased the remarkable panic provocation of the volunteers under 35% CO₂ inhalation stress [27]. However, anxiety and depression disorders are also enhanced in individuals with long-term opioid-related drug use [3–5]. These conflicting reports suggest that the role of MORs in modulating emotional responses may be bidirectional. MORs are extensively expressed in several emotion-related brain areas that control different aspects of the transmission of emotive information. The modulation of emotional behavior by MORs in different brain regions exhibit region-specificity. For example, MORs expressed on the basolateral amygdala and dorsal raphe nucleus induce anxiolytic effects upon their activation [29, 30], whereas MORs expressed on the central amygdala and lateral septum elicit anxiety-like behavior upon their activation in rodents [31, 32]. Thus, the precise mechanism by which different local brain regions of MORs modulate emotional responses is advantageous for risk aversion to emotional dysfunctions during MOR-targeted pain management. The PAG is a critical brainstem region that controls the activation of the descending pain inhibitory pathway. MORs are densely expressed in the PAG [7]. As shown in rodent studies, both endogenous and exogenous activation of these MORs induce remarkable analgesia [7, 33]. The PAG also efficiently participates in the modulation of emotional responses such as aversive and defensive behaviors [8–10]. Homogenetically activating PAG vglut2-positive excitatory neurons elicits anxiety-like behaviors in open field and light/dark tests [11]. Here, we found that activated PAG MORs caused a decrease in the percentage of open arm entries and time spent in the open arms by mice; moreover, these effects were reversed by pharmacological blockage of PAG MORs. Thus, PAG MORs have a functional role in modulating anxiety-related behavior, which may explain the elevated emotional disorders of patients using opioid-related drugs.

The most conspicuous results of our work were that the anxiety-like behaviors elicited by the activation of PAG MORs were primarily contributed by PAG MOR_Astro, whereas MOR_Glut and MOR_GABA only had a negligible effect on anxiety-related responses. Several morphological studies have confirmed that MORs are distributed in various types of neurocytes [7, 14]. However, their functional modulatory role in the CNS is not fully understood. By developing three lines of MOR-knockdown mice, including MOR_Glut−/−, MOR_GABA−/−, and MOR_Astro−/− [7, 15], we detected the functional role of divergent cell types of PAG MORs on the modulation of anxiety-like behaviors. We found that anxiety-like behaviors induced by the local activation of vlPAG MORs were reversed by the conditional knockdown of MOR_Astro. In contrast, anxiety-like behaviors were marginally modulated in MOR_Glut−/− or MOR_GABA−/− mice. These results are exhilarating because of confirmation of the functional role of PAG MOR_GABA in pain modulation. As an inhibitory G protein-coupled receptor (GiPCR), the activation of MOR_GABA could decrease GABAergic neuron excitability and subsequently weaken the tonic inhibition from GABAergic neurons to PAG-rostral ventromedial medulla (RVM) excitatory projections to elicit analgesic effects [7, 34]. Our results demonstrated the previously unappreciated role of MOR_GABA in modulating anxiety-related responses, with MOR_Astro displaying critical effects in these responses. Thus, theoretically, the targeted interception of the activation of MOR_Astro would reduce the aversive emotional responses without affecting the analgesic effects during PAG MOR activation. This would provide a potential possibility to address the elevated emotional disorders of patients using opioid-related drugs targeting astrocytes of the PAG. We also found that MOR_Glut slightly influenced the modulation of anxiety-like behaviors. This finding aligned with the observation that the chemogenetic activation of GiPCR signaling in vGlut2-positive neurons of the PAG minimally contributed to emotional responses [11]. Beyond the modulation of pain and emotion, the PAG is associated with several other brain-related functions, including respiratory, cardiovascular, and sleep and wakefulness regulation [35, 36]. Thus, MOR expression on PAG glutamatergic neurons might have a role in modulating autonomic nervous system function or sleep-wake reactions.

By using in vitro calcium imaging combined with chemical genetics [19, 20], we further demonstrated that the PAG MOR_Astro elicits anxiety-like behavior through an astrocytic calcium signaling mechanism. As confirmed previously, astrocytic GiPCR activation elevates calcium ion levels in astrocytes [18]. In the nucleus accumbens, the activation of MOR_Astro by DAMGO also elicited calcium oscillations in astrocytes [18]. In agreement with these studies, we found that the activation of PAG MORs induced significant enhancement of calcium levels in astrocytes, and this phenomenon was not blocked by TTX, indicating that this increase was specifically mediated by MOR_Astro activation. The chemical activation of calcium signaling in PAG astrocytes further confirmed that calcium signaling in PAG astrocytes efficiently induces anxiogenic behavior. In contrast, anxiety-like behaviors elicited by the activation of PAG MORs were reversed by specific interception of astrocytic calcium signaling. As reported earlier, astrocytic calcium signaling activation can promote the release of gliotransmitters and subsequently modulate neural excitability through the control of synaptic transmission [37–40]. Thus, the activation of MOR_Astro seemed to modulate synaptic transmission through astrocytic calcium signaling, thereby controlling the release of gliotransmitters in the PAG to elicit anxiogenic effects. However, further research is required to support our hypothesis.

Although our study has certain strengths, there are also some limitations. First, male mice alone were used because female mice show nominal differences in the MOR-dependent modulation of emotional responses compared to male mice. Additional investigations with female mice are required for improved understanding of the role of PAG MORs in the modulation of emotional responses. Second, we confirmed only the action of the MOR_Astro-induced astrocytic calcium signaling mechanism on the modulation of vlPAG MOR-induced anxiety-like behavior. However, the role of astrocytes in modulating brain behaviors is always accomplished by regulating neural activity through control of synaptic transmission or intrinsic excitability of neurons. Further electrophysiological studies should focus on explaining the astrocytic control of the neural mechanisms of PAG MOR-induced anxiogenic effects.

Ethical Approval

All animal experiments were carried out by following the guidelines of the Chinese Council on Animal Care. The study protocol was approved by the Animal Protection Committee of Shaanxi Normal University and the Animal Care Committee of The First Affiliated Hospital of Xi'an Medical University.

Competing interests

The authors declare that they have no conflict of interest.

Author contributions

Yinan Du and Zhiqiang Liu designed experiments, conceived the project and prepared the manuscript. Yinan Du performed behavior test, relevant surgery and in vitro calcium imaging. Aozhuo Zhang, Zhiwei Li and Yukui Zhao performed morphological test and statistical analysis. Other authors helped to write and refine the manuscript.

Fundings

This work was supported by the National Natural Science Foundation of China (NSFC) (grants 91949105, 82071516) and the Science and Technology Innovation Team Project of Xi'an Medical University (grants 2021TD14).

Cuitavi J, Torres-Pérez JV, Lorente JD, Campos-Jurado Y, Andrés-Herrera P, Polache A, et al. Crosstalk between mu-opioid receptors and neuroinflammation: Consequences for drug addiction and pain. Neurosci Biobehav Rev. 2023;145:105011.
Corder G, Castro DC, Bruchas MR, Scherrer G. Endogenous and exogenous opioids in pain. Annu Rev Neurosci. 2018;41:453–73.
Nummenmaa L, Karjalainen T, Isojärvi J, Kantonen T, Tuisku J, Kaasinen V, et al. Lowered endogenous mu-opioid receptor availability in subclinical depression and anxiety. Neuropsychopharmacology. 2020;45(11):1953–9.
Rogers AHMA, Kauffman BYMA, Bakhshaie JMD, McHugh RK, PhD, Ditre JW, PhD, Zvolensky MJ. PhD. Anxiety sensitivity and opioid misuse among opioid-using adults with chronic pain. Am J Drug Alcohol Abuse. 2019;45(5):470–478.
Rogers AH, Garey L, Raines AM, Allan NP, Schmidt NB, Zvolensky MJ. Anxiety sensitivity and opioid use motives among adults with chronic low back pain. Exp Clin Psychopharmacol. 2022;30(1):23–30.
Pellissier LP, Gandía J, Laboute T, Becker JAJ, Le Merrer J. Mu opioid receptor, social behaviour and autism spectrum disorder: reward matters. Br J Pharmacol. 2018;175(14):2750–69.
Du Y, Yu K, Yan C, Wei C, Zheng Q, Qiao Y et al. The contributions of mu-opioid receptors on glutamatergic and GABAergic neurons to analgesia induced by various stress intensities. eNeuro. 2022;9(3).
Silva C, McNaughton N. Are periaqueductal gray and dorsal raphe the foundation of appetitive and aversive control? A comprehensive review. Prog Neurobiol. 2019;177:33–72.
Tovote P, Esposito MS, Botta P, Chaudun F, Fadok JP, Markovic M, et al. Midbrain circuits for defensive behaviour. Nature. 2016;534(7606):206–12.
Reis FMCV, Mobbs D, Canteras NS, Adhikari A. Orchestration of innate and conditioned defensive actions by the periaqueductal gray. Neuropharmacology. 2023;228:109458.
Taylor NE, Pei J, Zhang J, Vlasov KY, Davis T, Taylor E et al. The role of glutamatergic and dopaminergic neurons in the periaqueductal gray/dorsal raphe: separating analgesia and anxiety. eNeuro. 2019;6(1).
Graeff FG, Silveira MC, Nogueira RL, Audi EA, Oliveira RM. Role of the amygdala and periaqueductal gray in anxiety and panic. Behav Brain Res. 1993;58(1–2):123–31.
Nguyen E, Chiang MC, Nguyen C, Ross SE. Brainstem modulation of nociception by periaqueductal gray neurons expressing the µ-opioid receptor in mice. Anesthesiology. 2023;139(4):462–75.
Won W, Kim D, Shin E, Lee CJ. Mapping astrocytic and neuronal µ-opioid receptor expression in various brain regions using MOR-mCherry reporter mouse. Exp Neurobiol. 2023;32(6):395–409.
Shi MM, Fan KM, Qiao YN, Xu JH, Qiu LJ, Li X, et al. Hippocampal mu-opioid receptors on GABAergic neurons mediate stress-induced impairment of memory retrieval. Mol Psychiatry. 2020;25(5):977–92.
Nam MH, Han KS, Lee J, Won W, Koh W, Bae JY, et al. Activation of astrocytic mu-opioid receptor causes conditioned place preference. Cell Rep. 2019;28(5):1154–e11665.
Shen W, Chen S, Xiang Y, Yao Z, Chen Z, Wu X, et al. Astroglial adrenoreceptors modulate synaptic transmission and contextual fear memory formation in dentate gyrus. Neurochem Int. 2021;143:104942.
Corkrum M, Rothwell PE, Thomas MJ, Kofuji P, Araque A. Opioid-mediated astrocyte-neuron signaling in the nucleus accumbens. Cells. 2019;8(6):586.
Srinivasan R, Lu TY, Chai H, Xu J, Huang BS, Golshani P, et al. New transgenic mouse lines for selectively targeting astrocytes and studying calcium signals in astrocyte processes in situ and in vivo. Neuron. 2016;92(6):1181–95.
Noh K, Cho WH, Lee BH, Kim DW, Kim YS, Park K, et al. Cortical astrocytes modulate dominance behavior in male mice by regulating synaptic excitatory and inhibitory balance. Nat Neurosci. 2023;26(9):1541–54.
González-Arias C, Sánchez-Ruiz A, Esparza J, Sánchez-Puelles C, Arancibia L, Ramírez-Franco J, et al. Dysfunctional serotonergic neuron-astrocyte signaling in depressive-like states. Mol Psychiatry. 2023;28(9):3856–73.
Bossuyt J, Van Den Herrewegen Y, Nestor L, Buckinx A, De Bundel D, Smolders I. Chemogenetic modulation of astrocytes and microglia: State-of-the-art and implications in neuroscience. Glia. 2023;71(9):2071–95.
Yu X, Taylor AMW, Nagai J, Golshani P, Evans CJ, Coppola G, et al. Reducing astrocyte calcium signaling in vivo alters striatal microcircuits and causes repetitive behavior. Neuron. 2018;99(6):1170–e11879.
Yu X, Moye SL, Khakh BS. Local and CNS-wide astrocyte intracellular calcium signaling attenuation in vivo with CalEx^fl°x mice. J Neurosci. 2021;41(21):4556–74.
Tavakoli NS, Malone SG, Anderson TL, Neeley RE, Asadipooya A, Bardo MT, et al. Astrocyte Ca2 + in the dorsal striatum suppresses neuronal activity to oppose cue-induced reinstatement of cocaine seeking. Front Cell Neurosci. 2024;18:1347491.
Kennedy SE, Koeppe RA, Young EA, Zubieta JK. Dysregulation of endogenous opioid emotion regulation circuitry in major depression in women. Arch Gen Psychiatry. 2006;63(11):1199–208.
Esquivel G, Fernández-Torre O, Schruers KR, Wijnhoven LL, Griez EJ. The effects of opioid receptor blockade on experimental panic provocation with CO2. J Psychopharmacol. 2009;23(8):975–8.
Pickar D, Cohen MR, Naber D, Cohen RM. Clinical studies of the endogenous opioid system. Biol Psychiatry. 1982;17(11):1243–76.
Sasaki K, Fan LW, Tien LT, Ma T, Loh HH, Ho IK. The interaction of morphine and gamma-aminobutyric acid (GABA)ergic systems in anxiolytic behavior: using mu-opioid receptor knockout mice. Brain Res Bull. 2002;57(5):689–94.
Welsch L, Colantonio E, Frison M, Johnson DA, McClain SP, Mathis V, et al. Mu opioid receptor-expressing neurons in the dorsal raphe nucleus are involved in reward processing and affective behaviors. Biol Psychiatry. 2023;94(11):842–51.
Hebb AL, Zacharko RM, Gauthier M, Trudel F, Laforest S, Drolet G. Brief exposure to predator odor and resultant anxiety enhances mesocorticolimbic activity and enkephalin expression in CD-1 mice. Eur J Neurosci. 2004;20(9):2415–29.
Le Merrer J, Cagniard B, Cazala P. Modulation of anxiety by mu-opioid receptors of the lateral septal region in mice. Pharmacol Biochem Behav. 2006;83(3):465–79.
Li C, Sugam JA, Lowery-Gionta EG, McElligott ZA, McCall NM, Lopez AJ, et al. Mu Opioid Receptor Modulation of Dopamine Neurons in the Periaqueductal Gray/Dorsal Raphe: A Role in Regulation of Pain. Neuropsychopharmacology. 2016;41(8):2122–32.
Bagley EE, Ingram SL. Endogenous opioid peptides in the descending pain modulatory circuit. Neuropharmacology. 2020;173:108131.
Zhang H, Zhu Z, Ma WX, Kong LX, Yuan PC, Bu LF, et al. The contribution of periaqueductal gray in the regulation of physiological and pathological behaviors. Front Neurosci. 2024;18:1380171.
Deng H, Xiao X, Wang Z. Periaqueductal Gray Neuronal Activities Underlie Different Aspects of Defensive Behaviors. J Neurosci. 2016;36(29):7580–8.
Bazargani N, Attwell D. Astrocyte calcium signaling: the third wave. Nat Neurosci. 2016;19(2):182–9.
Goenaga J, Araque A, Kofuji P, Herrera Moro Chao D. Calcium signaling in astrocytes and gliotransmitter release. Front Synaptic Neurosci. 2023;15:1138577.
Kofuji P, Araque A. G-protein-coupled receptors in astrocyte-neuron communication. Neuroscience. 2021;456:71–84.
Newman EA. Glial cell regulation of neuronal activity and blood flow in the retina by release of gliotransmitters. Philos Trans R Soc Lond B Biol Sci. 2015;370(1672):20140195.

No competing interests reported.

Download PDF