Contributions of dorsal medial prefrontal cortex GABAergic mu opioid receptors on acute restraint stress-induced anxiety-like behavior

Abstract

The body’s defensive emotional reaction to stress presents as anxiety. Nevertheless, long-term anxious stimuli can cause pathological anxiety disorders together with endocrine system, neural system, and immune system dysfunctions. Thus, researchers should understand the neural processes that govern anxiety during stress to create preventive and care plans. By combining behavioral tests, chemogenetic strategies, c-fos stainings, and whole-cell patch clamp recordings, this study looked into the function of dorsal medial prefrontal cortex (dmPFC) mu opioid receptors (MORs), specifically GABAergic MORs (MOR_GABA), in anxiety-like behavior brought on by acute restraint stress (ARS). The dmPFC MOR_GABA is triggered during ARS. This activation reduced the inhibitory synaptic input onto projection neurons that target the basolateral amygdala (BLA) by suppressing the activity of neighboring dmPFC GABAergic interneurons. In turn, mice that showed anxiety-like behavior also had higher neural activity of dmPFC-BLA projection neurons. When combined, these results describe a local, MOR_GABA-mediated dmPFC pathway that governs acute stress-induced anxiety reactions. These findings could offer a possible logical foundation for dmPFC-focused medication screening and anxiety treatment.

Introduction

Anxiety is a combative mental response that arises in reaction to stressful situations, namely injury, severe workloads, or threat. Acute anxiety reactions can help people adjust to abrupt fluctuations in their environment [1]. However, when activated repeatedly or over an extended period of time, these reactions progress to pathological anxiety disorders. Such conditions are related to severe dysfunctions of endocrine, neurological, and even immune systems [2, 3]. Thus, developing preventative and curative measures requires an understanding the neural mechanisms governing anxiety reactions during stress.

In the central nervous system (CNS), the medial prefrontal cortex (mPFC) involves the modulation of anxiety-related behavior in response to stress [4]. Per morphological and biochemical research, mice exposed to stressors show raised expression of neuronal activity markers, including c-fos (an immediate early gene), and promote glutamate and dopamine (excitatory neurotransmitters) release in the mPFC; therefore, the mPFC can participate in stress reaction control [5, 6]. Anatomically, the mPFC, specifically the dorsal mPFC (dmPFC) projects widely to the basolateral amygdala (BLA), lateral septum, and hypothalamus, all emotion-regulation areas [7–9]. Thus, stress-related impulses can be transmitted quickly and effectively. Researchers have highlighted the key role of mPFC in controlling emotions by showing that dmPFC-BLA projections are crucial for stress-related anxiety-like behavior [7].

Despite the well-established role of the dmPFC-BLA projection in stress-induced anxiety, the local neuromodulatory mechanisms, particularly those involving the endogenous opioid system (EOS) that govern this pathway remain poorly understood. Given that stress potently engages the EOS, elucidating the underlying cellular mechanisms is critical for understanding how the dmPFC-BLA pathway functions under stressful conditions [10–12]. These local molecular mechanisms may control anxiety-like behavior that is dependent on the dmPFC.

The EOS is a stress-triggered neuromodulatory system in the CNS [13]. It primarily consists of kappa (KORs), delta (DORs), and mu opioid receptors (MORs) [14]. Among them, MORs appear to be the most sensitive to stress [15, 16]. MORs are expressed not only in glutamatergic neurons (MOR_Glut) but also in GABAergic neurons (MOR_GABA) [17]. Through their properties as inhibitory G protein-coupled receptors (GPCRs), these receptors regulate synaptic transmission and are behaviorally involved in the modulation of stress responses [15–17]. Recent study clarified that MORs in the dmPFC govern the activity of neural networks during acute stress [17]. Thus, they can modify behavior attributed to stress-driven anxiety. To test this hypothesis, three mouse models that were selectively knocked down for MOR_Glut, MOR_GABA, and dmPFC whole MORs were created. Applying chemogenetic, morphological, and electrophysiological methods in conjunction with anxiety-like behavioral tests, dmPFC MOR helps in the regulation of anxiety-like actions brought on by acute restraint stress (ARS). Specifically, MOR_GABA, rather than MOR_Glut, controls the mentioned processes. Furthermore, ARS substantially triggers dmPFC-BLA projections, and this activation largely depends on MOR_GABA. Mechanistically, MOR_GABA blocked dmPFC-BLA projections that induce anxiety-like behavior during restraint stress by suppressing the activity of dmPFC inhibitory interneurons. Taken together, these studies highlight the physiological role of dmPFC MOR_GABA in controlling anxiety. This mechanistic understanding could theoretically support new diagnostic indicators and treatments for anxiety disorders.

Materials and methods

Animals

The present study involved 79 male C57BL/6J mice and 40 male Oprm1^flox/flox mice (both 8 to 14 weeks) following the Animal Protection Committee of Xi’an Medical University licenses, the Animal Protection Committee of Shaanxi Normal University licenses and the Chinese Animal Protection Commission guidelines. Female mice were not included in the experimental design to avoid estrous cycle-related variability in opioid responses. The Experimental Animal Center of Shaanxi Normal University MOE Key Laboratory of Modern Teaching Technology supplied all animals. Each separately vented cage contained five to six animals, with a 12-h light and dark cycle. The humidity was maintained at 60 ± 10%, and the temperature was 21 ± 1 ℃. From 8:00 to 12:00 A.M., all behavioral tests were conducted. To minimize individual variances, all animals had the opportunity to adjust to the surroundings for 5 min each day for 3 days.

ARS protocol

The ARS procedure has been explained previously [18]. Animals were kept in vented cylindrical plastic tubes (inner diameter: 30 mm; length of the straight part: 80 mm) under gentle restraint. Then, each restrained mouse spent 2 h in a separate ventilated enclosure to create stress. For each mouse, the ARS procedure was uniformly initiated at approximately 9:00 a.m. in a quiet, consistent environment with 22 ℃ and 20 lx. Animals in the unstressed (US) group, i.e., controls, were not physically restrained; instead, they spent 2 h in the same vented enclosures.

Emotional behavioral tests

Open-field tests (OFTs) and elevated plus maze (EPM) tests were conducted per previous research [19, 20]. The equipment for OFTs included four opaque walls (30 cm high) enclosing a 40 cm × 40 cm arena. For 10 min, each animal was left to roam freely in a corner that faced the walls. A video camera system was used to continually record the amount of time spent in the central region (20 × 20 cm) and the overall distance covered. Noldus Information Technology’s EthoVision software 1.9 (USA) was used for analysis. The EPM apparatus was 50 cm above the ground and had four arms: two open and two closed. These arms measured 30 cm in length and 7 cm in width. The open arms were left uncovered, whereas the closed arms were encased by opaque walls that were 16 cm high. All arms were connected by a central deck that measured 7 cm × 7 cm. For 5 min, each animal was left to roam freely after being positioned in the middle, facing an open arm. A video camera setup was used to capture the behavior, and Noldus Information Technology’s EthoVision software 1.9 (USA) helped analyze the recordings. The number of open-arm entries, time spent in the open arms, and overall distance covered were computed. The OFTs or EPM test equipment was carefully cleaned and disinfected with 75% ethyl alcohol after each emotional test.

Stereotaxic surgery and recombinant adeno-associated virus (rAAV) injection

Stereotaxic surgery and rAAV injection were conducted as previously described [17]. After using 4% isoflurane for anesthesia, the animals were placed in a stereotaxic brain device. Then, the mice constantly inhaled 1% isoflurane using a cannula to maintain anesthesia. For surgery, the scalp was cut and hair removed, exposing the skull. The following rAAVs were used as our previous described [17] (all 170 nL, injecting speed: 40 nL/min):

• rAAV2/9-CAG-CRE-Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element (WPRE)-hGH-pA (5.28 × 10¹² vg/mL; brain VTA, Wuhan, China).

• rAAV2/9-mDLx-CRE-WPRE-hGH-pA (5.28 × 10¹² vg/mL; brain VTA, Wuhan, China).

• rAAV2/9-VGLUT1-CRE-WPRE-hGH-pA (5.83 × 10¹² vg/mL; brain VTA, Wuhan, China).

The bilateral dmPFCs of Oprm1^flox/flox mice received these constructs (mediolateral [ML]:± 0.25 mm; anteroposterior [AP]: 1.75 mm; and dorsoventral [DV]: -2.50 mm) to create conditional MOR knockdown (cKD), MOR_Glut cKD, or MOR_GABA cKD models. C57BL/6J mice received parallel injections for their matching controls (MOR WT, MOR_Glut WT, or MOR_GABA WT).

The combined strategies were also applied in some experiments as follows:

1. rAAV injection + patch-clamp reading or c-Fos staining: The bilateral BLAs (mediolateral [ML]:± 3.00 mm; anteroposterior [AP]: -1.50 mm; and dorsoventral [DV]: -4.80 mm) of MOR_GABA WT and MOR_GABA cKD mice received rAAV2/Retro-Enhanced Green Fluorescent Protein (EGFP)-WPRE-pA (300 nL/injection, injecting speed: 40 nL/min, 1.00 × 10¹³ vg/mL, Taitool Bioscience, Shanghai, China). Each mouse was allowed to recover at least 4 weeks before patch-clamp readings or c-Fos staining.

2. rAAV injection + chemogenetic silencing: The bilateral BLAs of C57BL/6J mice were injected with rAAV2/Retro-hsyn-hM4D-EGFP-WPRE-pA (300 nL/injection, injecting speed: 40 nL/min, 1.00 × 10¹³ vg/mL, Taitool Bioscience, Shanghai, China). After 3 weeks, intracranial cannula were bilaterally inserted into the mPFCs as our previous studies described [17]. And the intra-mPFC injections were performed after 1 week of recovery. CNO (10 µM, 0.1 µL/injection per side, Taitool Bioscience, Shanghai, China) was delivered into the cannulas by means of circulating drug delivery system. Saline intra-mPFC administration was set at control. The ARS procedure or behavioral tests were proceeded for 30 min after CNO injection.

Fluorescence in situ hybridization

Using our previously published techniques [17], MOR deletion in the dmPFC of MOR_Glut cKD, MOR cKD, and MOR_GABA cKD mice was detected via fluorescence in situ hybridization [17, 21]. Pentobarbital sodium (50 mg/kg body weight) was injected intraperitoneally to induce profound anesthesia, and the animals were perfused for 5 min with 0.9% saline. At -22 ℃, Sakura Finetek’s optimal cutting temperature (OCT) compound (catalog no: Tissue-Tek 4583; Torrance, CA, USA) helped quickly harvest and implant the brains. Leica Biosystems’ CM1950 cryostat (Nussloch, Germany) was used to prepare mPFC in frozen slices (16 μm). At room temperature, these slices were dehydrated using a graded ethanol sequence (50%, 75%, and 100%), with each stage extending for 5 min, after being fixed at 4 °C in 4% paraformaldehyde (PFA) for 15 min. Then, these slices underwent 15 min of hydrogen peroxide treatment and 20 min of protease IV treatment. ACD’s Advanced Cell Diagnostics (Newark, CA, USA) provided the probes that target Vgat (catalog no: 424548-C3), Oprm1 (catalog no: 544731-C2), and Vglut1 (catalog no: 416631-C1). These probes were conjugated to Atto 650, Atto 570, and Atto 520, respectively. ACD’s HybEZ™ oven helped with in situ hybridization per the manufacturer’s directions for the RNAscope Multiplex Fluorescent Reagent Kit (catalog no: 323100; ACD). Finally, an anti-fade mounting medium containing DAPI was used to mount these slices; Zeiss’ fluorescence microscope (ObercKOchen, Germany) helped capture pictures. Image count and calculation were analyzed using ImageJ.

Immunofluorescence for c-fos staining

In MOR_GABA CKD and MOR_GABA WT mice, the c-fos immunofluorescence test was used to assess the activating impact of dmPFC-BLA projections (before and 1 h after restraint stress). The experiment followed the techniques stated before [17]. Briefly, sodium pentobarbital was used to induce profound anesthesia in all animals; then, they were transcardially perfused with 0.9% saline as well as 4% paraformaldehyde (added to 0.1 M phosphate-buffered saline (PBS)) in order. After being extracted and post-fixed for 24 h, the brains were cryoprotected by immersing them in 30% sucrose (added to 0.1 M PBS) for 48 h at 4 ℃. Next, 20 μm frozen slices comprising the mPFC were created after embedding the brain slices in OCT compound. These slices were initially allowed to adjust to room temperature before being blocked in PBS containing 0.5% Triton X-100 and 10% non-immune donkey serum for 60 min to minimize non-specific binding for immunofluorescence labeling. They were cultivated with Cell Signaling Technology’s rabbit anti-c-fos primary antibody (1:500; #2250; Danvers, MA, USA) for the entire night at 4 °C. The following day, these slices were incubated with Abcam’s Alexa Fluor 594-conjugated donkey anti-rabbit secondary antibody (1:1000; ab150076; Boston, MA, USA) for 2 h in a dark-proof environment after being rinsed with 0.1 M PBS thrice. Finally, after applying an anti-fade mounting medium to these sections, a fluorescence microscope helped capture fluorescence images. Image J was used to analyze the image’s count and calculation.

Electrophysiological recording

Alterations in synaptic transmission of dmPFC-BLA projections were investigated through electrophysiological recordings. The procedures were conducted as published earlier [22]. Mice were quickly decapitated after being sedated with 50 mg/kg pentobarbital sodium. Brains were extracted immediately after and placed at 4 ℃ in oxygenated adjusted artificial cerebrospinal fluid (sucrose-substituted ACSF). ACSF composition (in mM) was 2.5 KCl, 250 sucrose, 1.2 NaH₂PO₄, 1.2 MgCl₂, 26 NaHCO₃, 2.4 CaCl₂, and 11 glucose. Then, the brains were immersed in 4 °C ACSF after being placed on Vibratome Company’s vibrating slicer (1000 Plus; St. Louis, MO, USA). Next, 300-µm-thick coronal sections comprising the dmPFC were cut and cultivated at room temperature in oxygenated standard ACSF (pH 7.4) for 30 min. ACSF consisted of the following concentrations (in mM): 2.5 KCl, 125 NaCl, 2 MgCl₂, 1.25 NaH₂PO₄, 25 glucose, 2 CaCl₂, and 25 NaHCO₃. These slices were moved to an immersion recording container after incubation, where they were superfused with 2 mL/min oxygenated standard ACSF. Molecular Devices’ Multiclamp 700B amplifier (Sunnyvale, CA, USA) was used for whole-cell readings. EGFP-positive dmPFC-BLA projections were the focus of these readings. A DM LFSA upright microscope (Leica, Germany) fitted with a 488-nm fluorescent module and infrared differential interference contrast optics helped visualize these neurons. When loaded with the internal solution containing 60 CsCl, 10 HEPES, 80 CsCH₃SO₃, 2 MgCl₂, 2 QX-314, 4 MgATP, 0.2 EGTA, 0.3 Na₂GTP, and 10 Na₂-phosphocreatine (pH adjusted to 7.2 using CsOH) (in mM), the recording pipettes exhibited a resistance of 4 to 8 MΩ. To record mIPSCs, whole-cell patch-clamp experiments were conducted at a holding potential of -70 mV. The recording solution contained TTX (1 µM, Sigma-Aldrich, USA) to isolate miniature synaptic events, as well as AP5 (100 µM, Sigma, USA) and DNQX (100 µM, Sigma, USA) to block NMDA and AMPA receptors, thereby eliminating excitatory currents. Clampfit 10.5 was used for mIPSC analysis.

Statistical analysis

SPSS version 22.0 was used to analyze the data, which are displayed as mean ± standard error of the mean (S.E.M.), or as mean ± S.E.M. % (in some cases). The following tests were conducted: one-way analysis of variance (ANOVA), repeated ANOVA, two-way ANOVA, paired Student’s t test, and unpaired Student’s t test. After each ANOVA, Sidak’s multiple-comparisons tests were conducted. In every experiment, the significance level was chosen as p < 0.05. Figure legends elucidate the statistical data.

Results

ARS induces dmPFC MOR-dependent anxiety-like behavior

MORs modulate emotional responses [11, 23], but the mechanisms underlying dmPFC MORs remain poorly understood. To this end, the ARS protocol, a common paradigm of acute anxiety-like behavior [18], was used in both MOR WT and MOR cKD mice for emotional behavioral tests. First, the results affirmed that the ARS protocol successfully elicited anxiety-like behavior (Fig. 1A). As anticipated, animals that received ARS showed markedly stronger anxiety-like behavior, compared with the US group. Specifically, animals exposed to ARS spent a shorter span in the center zone of the OFTs (Figs. 1B, C). Moreover, in EPM tests, they spent a shorter period in the open arms and demonstrated fewer open arm entries (Figs. 1D, E). Notably, neither the OFT (Fig. 1C) nor the EPM tests (Fig. 1E) showed any significant difference in the overall distance covered between the US and ARS groups, suggesting that behavioral changes were attributed to anxiety rather than motor abnormalities. Collectively, ARS is a useful paradigm for eliciting acute anxiety-like behavior.

Fig. 1.

Acute restraint stress (ARS) induces anxiety-like behavior. (A) Experimental procedure. (B) Real-time activity monitoring in open field tests (OFTs). (C) Summary plots of the total distance traveled (unpaired Student’s t test, t₁₄ = 1.706, p = 0.1100) and period spent in the OFT center (unpaired Student’s t test, t₁₄ = 6.944, p < 0.0001 (**)). US group, n = 8 mice; ARS group, n = 8 mice. (D) Real-time activity monitoring in elevated plus maze (EPM) tests. (E) Summary plots of the total distance traveled (unpaired Student’s t test, t₁₄ = 1.682, p = 0.1148), duration used in the EPM open arms (unpaired Student’s t test, t₁₄ = 6.715, p < 0.0001 (**)) and its entries (unpaired Student’s t test, t₁₄ = 3.782, p = 0.0020 (**)). US group, n = 8 mice; ARS group, n = 8 mice

Next, the role of dmPFC MORs on modifying anxiety-like behavior was investigated. dmPFC MOR cKD and MOR WT mice were created as described earlier (Figs. 2A, D). The knockdown efficiency was validated by fluorescence in situ hybridization and subsequent cell counting tests, which confirmed significantly reduced MOR expression in the dmPFC of MOR cKO mice (Figs. 2B, C). By spending a shorter period in the center zone during OFTs (Fig. 2E) and a shorter span in the open arms, along with fewer open arm entries during EPM tests (Fig. 2F), MOR WT mice exposed to ARS demonstrated significant anxiety-like responses in behavioral tests. However, MOR cKD mice did not exhibit these ARS-mediated alterations (Figs. 2E, F). Furthermore, MOR cKD mice in both US and ARS groups showed comparably low levels of anxiety-like responses (Figs. 2E, F). Thus, MOR knockdown inhibited the expression of behavioral abnormalities produced by ARS. Additionally, neither the OFTs nor the EPM test results demonstrated any variations in the total distance covered between the groups (Figs. 2E, F), confirming that the observed behavioral effects were not attributable to altered locomotion. These results suggest the crucial function of dmPFC MORs in regulating ARS-mediated anxiety-like behavior.

Fig. 2.

Contributions of mu opioid receptors (MORs) in the dorsomedial prefrontal cortex (dmPFC) in controlling ARS-mediated anxiety-like behavior. (A) Schematic diagram of the generation of MOR conditional knockdown (cKD) mice and MOR wildtype (WT) mice (controls). (B) In situ hybridization for DAPI (blue) and Oprm1 mRNA (red) in dmPFC MOR cKD mice and MOR WT mice. Scale bar: 20 μm. Oprm1-positive cells are shown by white arrowheads. (C) Analyzing Oprm1-positive cell percentage against dmPFC cells (unpaired Student’s t test, t4 = 14.79, p < 0.0001 (**)). MOR cKD group: n = 3; MOR WT group: n = 3. (D) Experimental procedure. (E) Overall distance traveled (F(3,28) = 4.568, p = 0.0100 (**); MOR WT ARS vs. MOR WT US, p = 0.4798; MOR cKD ARS vs. MOR cKD US, p = 0.7475; MOR cKD ARS vs. MOR WT ARS, p = 0.6543; one-way ANOVA and Šidák correction) and time spent in the center area (F(3,28) = 10.09, p = 0.0001 (**); MOR WT ARS vs. MOR WT US, p = 0.0001 (**); MOR cKD ARS vs. MOR cKD US, p = 0.9999; MOR cKD ARS vs. MOR WT ARS, p = 0.0031 (**); one-way ANOVA and Šidák correction) of OFTs. MOR WT ARS group: n = 8; MOR WT US group: n = 8; MOR cKD ARS group: n = 8; and MOR cKD US group: n = 8. (F) Overall distance traveled (F(3,28) = 0.9563, p = 0.4270, MOR WT ARS vs. MOR WT US, p = 0.6772; MOR cKD ARS vs. MOR cKD US, p = 0.9983; MOR cKD ARS vs. MOR WT ARS, p = 0.8783; one-way ANOVA and Šidák correction), open arm duration (F(3,28) = 12.56, p < 0.0001; MOR WT ARS vs. MOR WT US, p < 0.0001(**); MOR cKD ARS vs. MOR cKD US, p = 0.9745; MOR cKD ARS vs. MOR cKD ARS, p = 0.0067 (**); one-way ANOVA and Šidák correction) and open arm entries (F(3,28) = 16.51, p < 0.0001 (**); MOR WT ARS vs. MOR WT US, p < 0.0001 (**); MOR cKD ARS vs. MOR cKD US, p = 0.9938; MOR cKD ARS vs. MOR WT ARS, p < 0.0001 (**); one-way ANOVA and Šidák correction) in EPM tests. MOR WT ARS group: n = 8; MOR WT US group: n = 8; MOR cKD ARS group: n = 8; and MOR cKD US group: n = 8

dmPFC MOR_GABA, but not MOR_Glut, modulates anxiety-like behavior

The mentioned results demonstrated that dmPFC MORs have a functional role in regulating ARS-induced acute anxiety-like reactions. However, MORs are expressed in both excitatory glutamatergic neurons and inhibitory GABAergic neurons in the dmPFC [24]. Clarifying the mechanism by which dmPFC MORs contribute to regulating anxiety-like behavior is therefore warranted (Figs. 3D and 4D). To this end, MOR_GABA cKD and dmPFC MOR_Glut cKD mice, along with their corresponding wild-type controls (MOR_GABA WT and MOR_Glut WT mice, respectively) were created (Figs. 3A and 4A). Fluorescence in situ hybridization results validated the successful knockdown of dmPFC MOR_Glut in MOR_Glut cKD mice (Figs. 3B, C) and of dmPFC MOR_GABA in MOR_GABA cKD mice (Figs. 4B, C). Next, these models helped confirm the role of MOR_Glut and MOR_GABA in regulating anxiety-like behavior.

Fig. 3.

dmPFC glutamatergic MORs (MOR_Glut) do not control ARS-induced anxiety-like behavior. (A) Schematic diagram of the generation of MORGlut cKD and MORGlut WT mice. (B) In situ hybridization for vglut1 mRNA (green), DAPI (blue), and Oprm1 mRNA (red) in the dmPFC in MORGlut cKD and MORGlut WT mice. Scale bar: 20 μm. Double-labeled cells with vglut1 and Oprm1 mRNA are shown by white arrowheads. (C) Double-positive cell percentage (vglut1 and Oprm1) against vglut1 positive cells. (unpaired Student’s t test, t4 = 25.93, p < 0.0001(**)). MORGlut cKD group: n = 3; MORGlut WT group: n = 3. (D) Experimental procedure. (E) Overall distance (F(3,28) = 0.9062, p = 0.4505; MORGlut WT ARS vs. MORGlut WT US, p > 0.9999; MORGlut cKD ARS vs. MORGlut cKD US, p = 0.9886; MORGlut cKD ARS vs. MORGlut WT ARS, p = 0.9923; one-way ANOVA and Šidák correction) and period spent in the OFT center (F(3,28) = 14.85, p < 0.0001 (**); MORGlut WT ARS vs. MORGlut WT US, p = 0.0003 (**); MORGlut cKD ARS vs. MORGlut cKD US, p = 0.0010 (**); MORGlut cKD ARS vs. MORGlut WT ARS, p = 0.6685; one-way ANOVA and Šidák correction). MORGlut WT ARS group: n = 8; MORGlut WT US group: n = 8; MORGlut cKD ARS group: n = 8; and MORGlu cKD US group: n = 8. (F) Overall distance (F(3,28) = 1.353, p = 0.2774, MORGlut WT ARS vs. MORGlut WT US, p = 0.3590; MORGlut cKD ARS vs. MORGlut cKD US, p = 0.9863; MORGlut cKD ARS vs. MORGlut WT ARS, p = 0.9994; one-way ANOVA and Šidák correction), period spent in EPM open arms (F(3,28) = 17.35, p < 0.0001 (**); MORGlut WT ARS vs. MORGlut WT US, p = 0.0002 (**); MORGlut cKD ARS vs. MORGlut cKD US, p = 0.0005 (**); MORGlut cKD ARS vs. MORGlut WT ARS, p = 0.3420; one-way ANOVA and Šidák correction) and open arm entries (F(3,28) = 6.959, p = 0.0012 (**); MORGlut WT ARS vs. MORGlut WT US, p = 0.0127 (*); MORGlut cKD ARS vs. MORGlut cKD US, p = 0.0457 (*); MORGlut cKD ARS vs. MORGlut WT ARS, p = 0.9965; one-way ANOVA and Šidák correction). MORGlut WT ARS group: n = 8; MORGlut WT US group: n = 8; MORGlut cKD ARS group: n = 8; and MORGlu cKD US group: n = 8

Fig. 4.

Key contributions of dmPFC gamma-aminobutyric acid-ergic MORs (MOR_GABA) on ARS induced anxiety-like behavior. (A) Schematic diagram of the generation of MORGABA cKD and MORGABA WT mice. (B) In situ hybridization for DAPI (blue), vGAT mRNA (green), and Oprm1 mRNA (red) in the dmPFC in MORGABA WT mice and MORGABA cKD mice. Scale bar: 20 μm. Double-labeled with vGAT and Oprm1 mRNA are shown by white arrowheads. (C) Double-positive cell percentage (vGAT and Oprm1) against vGAT-positive cells. (unpaired Student’s t test, t4 = 25.93, p < 0.0001 (**)). MORGABA cKD group: n = 3; MORGABA WT group: n = 3. (D) Experimental procedures. (E) Overall distance traveled (F(3,28) = 1.688, p = 0.1992, MORGABA WT ARS vs. MORGABA WT US, p > 0.9999; MORGABA cKD ARS vs. MORGABA cKD US, p = 0.2857; MORGABA cKD ARS vs. MOR WT ARS, p = 0.9995; one-way ANOVA and Šidák correction) and duration spent in OFT center (F(3,28) = 21.86, p < 0.0001 (**); MORGABA WT ARS vs. MORGABA WT US, p < 0.0001 (**); MORGABA cKD ARS vs. MORGABA cKD US, p = 0.0698; MORGABA cKD ARS vs. MORGABA WT ARS, p = 0.0002 (**); one-way ANOVA and Šidák correction). MORGABA WT ARS group: n = 8; MORGABA WT US group: n = 8; MORGABA cKD ARS group: n = 8; and MORGABA cKD US group: n = 8. (F) Overall distance traveled (F(3,28) = 2.184, p = 0.1122, MORGABA WT ARS vs. MORGABA WT US, p > 0.9999, MORGABA cKD ARS vs. MORGABA cKD US, p = 0.8563; MOR cKD ARS vs. MOR WT ARS, p = 0.1711; one-way ANOVA and Šidák correction), duration spent in EPM open arms (F(3,28) = 15.60, p < 0.0001 (**); MORGABA WT ARS vs. MORGABA WT US, p < 0.0001 (**); MORGABA cKD ARS vs. MORGABA cKD US, p = 0.8867; MORGABA cKD ARS vs. MORGABA WT AR, p = 0.0017 (**); one-way ANOVA and Šidák correction) and open arm entries (F(3,28) = 8.343, p = 0.0004 (**); MORGABA WT ARS vs. MORGABA WT US, p = 0.0005 (**); MORGABA cKD ARS vs. MORGABA cKD US, p = 0.9635; MORGABA cKD ARS vs. MORGABA WT ARS, p = 0.0209 (*); one-way ANOVA and Šidák correction). MORGABA WT ARS group: n = 8; MORGABA WT US group: n = 8; MORGABA cKD ARS group: n = 8; and MORGABA cKD US group: n = 8

Both MOR_Glut cKD mice and their WT controls showed markedly increased anxiety-like responses after being exposed to ARS behavioral tests. Moreover, they spent significantly shorter periods in the central area during OFTs (Fig. 3E). Additionally, MOR_Glut cKD and MOR_Glut WT mice spent a shorter span in the open arms during EPM tests, besides exhibiting fewer open arm entries (Fig. 3F), compared with the corresponding US group. These results suggest a non-essential role of MOR_Glut in regulating ARS-dependent anxiety-like behavior. This is because MOR knockdown in glutamatergic neurons did not change ARS-induced anxiety tendencies.

For MOR_GABA mice, however, the results were notably different. Compared with their US controls, only MOR_GABA WT mice that received ARS exhibited significant anxiety-like behavior, as confirmed by the shorter span spent in the OFTs center zone (Fig. 4E), along with less time spent in the EPM open arms and fewer open arm entries (Fig. 4F). Interestingly, MOR_GABA cKD US mice and MOR_GABA cKD mice that received ARS showed comparable emotional responses in both OTFs and EPM tests (Figs. 4E, F). Thus, dmPFC MOR_GABA is central to modulating ARS-triggered anxiety-like behavior. Locomotor activity did not differ among groups in the OFT or EPM test (Figs. 3E and F and 4E and F), confirming behavioral assessment reliability.

dmPFC MOR_GABA modulates activity of dmPFC-BLA projections during anxiety-like behavior

Anxiety-like behavior is predominantly regulated by dmPFC MOR_GABA under ARS exposure. The dmPFC sends numerous excitatory projections to the BLA, which govern the production and maintenance of anxiety-like behavior in response to acute stress [7, 10]. It is well proofed that optogenetic stimulation of the dmPFC-BLA pathway promoted enhanced anxiety-related behaviors in rodents [7]. Thus, it was hypothesized that MOR_GABA-mediated regulation of dmPFC-BLA projection activity is vital for ARS-induced anxiety. Behavioral tests were conducted in conjunction with c-Fos staining and chemogenetic techniques to affirm these phenomena.

First, whether ARS-induced anxiety-like behavior necessitates increased activity of dmPFC-BLA projections was evaluated (Figs. 5A-C). As described earlier [25, 26], an inhibitory designer receptors exclusively activated by designer drugs (DREADDs) approach under ARS was used to specifically inhibit these projections (Figs. 5A, B). Animals received intracranial cannula injections of either saline as a control (hM4Di + SAL group) or clozapine-N-oxide (CNO) to suppress dmPFC–BLA projections after 28 days of viral expression as our previous study described [17].

Fig. 5.

Inhibition of dmPFC-basolateral amygdala (BLA) projections changes ARS-induced anxiety-like behavior. (A) dmPFC-BLA projection labeling by enhanced green fluorescent protein (EGFP). (B) Labeled fluorescence images showing injecting site (BLA) and tracing site (dmPFC). Scale bar: 100 μm. (C) Experimental procedures. (D) Overall distance traveled (F (3,28) = 0.1628, p = 0.9205, hM4Di CNO US vs. hM4Di saline US, p > 0.9999; hM4Di saline ARS vs. hM4Di saline US, p = 0.9893, hM4Di CNO ARS vs. hM4Di CNO US, p > 0.9999; hM4Di CNO ARS vs. hM4Di saline ARS, p = 0.9953; one-way ANOVA and Šidák correction) and duration spent in OFT center area (F(3,28) = 9.851, p = 0.0001 (**); hM4Di CNO US vs. hM4Di saline US, p = 0.3441; hM4Di saline ARS vs. hM4Di saline US, p = 0.0165 (*); hM4Di CNO ARS vs. hM4Di CNO US, p = 0.8151; hM4Di CNO ARS vs. hM4Di saline ARS, p = 0.0025 (**); one-way ANOVA and Šidák correction) of OFTs. hM4Di saline ARS group: n = 8; hM4Di saline US group: n = 8; hM4Di CNO ARS group: n = 8; and hM4Di CNO US group: n = 8. (E) Overall distance traveled (F(3,28) = 0.5148, p = 0.6754; hM4Di CNO US vs. hM4Di saline US, p = 0.9997; hM4Di saline ARS vs. hM4Di saline US, p = 0.9475; hM4Di CNO ARS vs. hM4Di CNO US, p = 0.9703; hM4Di CNO ARS vs. hM4Di saline ARS; p > 0.9999; one-way ANOVA and Šidák correction), duration spent in EPM open arms (F(3,28) = 17.04, p < 0.0001 (**); hM4Di CNO US vs. hM4Di saline US, p = 0.9117; hM4Di saline ARS vs. hM4Di saline US, p < 0.0001 (**); hM4Di CNO ARS vs. hM4Di CNO US, p = 0.6839; hM4Di CNO ARS vs. hM4Di saline ARS, p = 0.0001(**); one-way ANOVA and Šidák correction) and open arm entries (F(3,28) = 9.796, p = 0.0001 (**); hM4Di CNO US vs. hM4Di saline US, p = 0.9995; hM4Di saline ARS vs. hM4Di saline US, p = 0.0006 (**); hM4Di CNO ARS vs. hM4Di CNO US, p = 0.9995; hM4Di CNO ARS vs. hM4Di saline ARS, p = 0.0006 (**); one-way ANOVA and Šidák correction). hM4Di saline ARS group: n = 8; hM4Di saline US group: n = 8; hM4Di CNO ARS group: n = 8; and hM4Di CNO US group: n = 8

There was an insignificant difference in emotional performances between the hM4Di + CNO US and hM4Di + SAL US groups in both OFTs and EPM tests (Figs. 5D, E). Thus, the basal activity of dmPFC-BLA projections plays a marginal role in modulating anxiety-like behavior. Contrarily, mice in the hM4Di + CNO ARS group spent a markedly longer period in the OFTs center area than mice in the hM4Di + SAL ARS group (Fig. 5D). Moreover, they spent a longer span in the EPM open arms and demonstrated higher number of open arm entries, compared with mice in the hM4Di + SAL ARS group (Fig. 5E). Mice in both hM4Di + CNO US and hM4Di + CNO ARS groups demonstrated no significant anxiety-like responses in both OFTs and EPM tests (Figs. 5D, E). Therefore, ARS-dependent anxiety-like behavior depends on dmPFC-BLA projection activity. Locomotor activity was nondistinctive across groups in both tests (Fig. 5E and F), confirming behavioral reliability.

Next, we determined whether the dmPFC MOR_GABA controls dmPFC-BLA projection activity. c-Fos expression was evaluated in dmPFC-BLA projections to assess this result. Both, MOR_GABA cKD and MOR_GABA WT mice demonstrated fluorescently tagged dmPFC-BLA projections (Fig. 6A). Moreover, c-Fos expression was detected during both US and ARS exposure (Fig. 6B). Compared with US controls, ARS significantly enhanced c-Fos immunoreactivity within tagged dmPFC-BLA projections in MOR_GABA WT mice (Fig. 6B). In contrast, MOR_GABA cKD mice exhibited limited c-fos immunoreactivity in these projections in both conditions (Fig. 6B). These discrepancies were further validated by a quantitative analysis (Fig. 6C). Collectively, dmPFC MOR_GABA modulates dmPFC-BLA projection activity during ARS exposure.

Fig. 6.

dmPFC MOR_GABA modulates dmPFC-BLA projection activity under 2 h of ARS. (A) dmPFC-BLA projection labeling in MORGABA WT mice and MORGABA cKD mice for c-fos staining. (B) c-fos (red) expression (in BLA-projecting dmPFC neurons marked by EGFP). Scale bar: 50 μm. EGFP and c-fos co-labeling is shown by white arrowheads. (C) c-fos (+)-labeled dmPF-BLA projection percentage in MORGABA cKD US and MORGABA cKD ARS mice (top, t4 = 2.328, p = 0.0804, unpaired Student’s t test), and that in MORGABA WT US and MORGABA WT ARS mice (bottom, t4 = 3.532, p = 0.0242 (*), unpaired Student’s t test). MORGABA WT ARS group: n = 3; MORGABA WT US group: n = 3; MORGABA cKD ARS group: n = 3; and MORGABA cKD US group: n = 3

MOR_GABA modulates dmPFC-BLA projection activity by decreasing presynaptic GABAergic neurotransmission

Through specific “disinhibition” mechanisms, MOR_GABA increases the activity of dmPFC pyramidal projections [17]. Thus, it was hypothesized that dmPFC MOR_GABA may regulate anxiety-like behavior after ARS exposure through comparable pathways. To test this hypothesis, whole-cell patch-clamp recordings were used to identify GABAergic inputs onto dmPFC-BLA projections (Figs. 7A, B). The frequency of mIPSCs decreased after ARS exposure, but not their amplitude, which is in line with earlier observations. Upon comparing MOR_GABA WT ARS mice to MOR_GABA WT US mice, mIPSC events demonstrated a decrease in the normalized mIPSC frequency (Fig. 7C, D), along with a shift to the right in the cumulative frequency distribution (Figs. 7C, D). Therefore, ARS exposure reduces presynaptic GABAergic inputs onto dmPFC-BLA projections. However, MOR_GABA cKD mice did not exhibit these alterations. Neither the normalized mIPSC frequency (Figs. 7C, D) nor the cumulative frequency distribution of mIPSC events (Figs. 7C, D) demonstrated any significant difference between MOR_GABA cKD ARS mice and MOR_GABA cKD US mice. We also found no significant difference in mIPSC frequencies between the MOR_GABA WT US group and the MOR_GABA cKD US group, indicating that MOR_GABA lacks tonic inhibition at baseline but is selectively engaged under stress. Furthermore, the mIPSC frequency and amplitude did not differ between the MOR_GABA WT US and MOR_GABA cKD US groups (Figs. 7C-F), suggesting that MOR_GABA did not affect basal GABAergic transmission.

Fig. 7.

dmPFC MOR_GABA attenuates inhibitory synaptic transmission on dmPFC-BLA projections under 2 h of ARS. (A) dmPFC-BLA projection labeling in MORGABA WT mice and MORGABA cKD mice for whole cell patch clamp. (B) Schematic diagram of mPFC-BLA projection labeling in MORGABA WT mice and MORGABA cKD mice. (C) dmPFC-BLA projections showing miniature inhibitory postsynaptic currents (mIPSCs). (D) Left, mean mIPSC frequency in dmPFC-BLA projections (F(3,28) = 35.79, p < 0.0001 (**); MORGABA WT ARS vs. MORGABA WT US; p < 0.0001 (**); MORGABA cKD ARS vs. MORGABA cKD US, p = 0.8283; MORGABA cKD ARS vs. MORGABA WT ARS, p < 0.0001 (**); one-way ANOVA and Šidák correction). MORGABA WT ARS group: n = 8 cells from n = 2 mice; MORGABA WT US group: n = 8 cells from n = 2 mice; MORGABA cKD ARS group: n = 8 cells from n = 3 mice; and MORGABA cKD US group: n = 8 cells from n = 2 mice. Right, cumulative probability of the inter-event intervals from left. (E) Left, mean mIPSC amplitude in dmPFC-BLA projections. (F(3,28) = 0.2395, p = 0.8680; MORGABA WT ARS vs. MORGABA WT US; p = 0.9905; MORGABA cKD ARS vs. MORGABA cKD US, p = 0.9957; MORGABA cKD ARS vs. MORGABA WT ARS, p = 0.9783; one-way ANOVA and Šidák correction). Right, cumulative probability of the amplitude from left

Taken together, these findings highlight the importance of dmPFC MOR_GABA in regulating ARS-dependent anxiety-like behavior. This effect might be attributed to MOR_GABA-mediated improvements in dmPFC-BLA projection activity through presynaptic GABAergic neurotransmission.

Discussion

The results demonstrated that dmPFC MOR_GABA substantially regulates anxiety-like behavior after ARS exposure. MOR_GABA enhances the functional activity of dmPFC-BLA projections by decreasing inhibitory inputs, which causes elevated anxiety-like reactions in behavioral tests conducted under ARS. These results provide a specific dmPFC MOR_GABA-mediated circuit mechanism underscoring anxiety brought on by acute stress.

Pain, composure, addiction, and fear memory are some behavioral traits largely modulated by MORs as a prototypical inhibitory GPCR [23, 27, 28]. Researchers have demonstrated the mechanism by which opioid receptors control emotional reactions [21, 29, 30]. For example, GABAergic plasticity induced by MOR activation in the striatum promotes negative affective states [29]. Likewise, MORs in the habenula drive unpleasant effects in a project-specific manner [30]. However, the precise function of cortical MORs in controlling emotional responses remains elusive. It is well established that the mPFC is critically involved in the regulation of emotional responses [7, 19]. Notably, the dmPFC, which encompassing most of the prelimbic (PrL) area and a dorsal part of the infralimbic (IL) area, sends extensive excitatory projections to the BLA [7, 19, 31]. These projections are positively activated under stress stimulation and contribute to the induction of anxiety-like behaviors [7, 19]. Morphological studies have revealed abundant MOR expression in the dmPFC, which is strongly implicated in the regulation of emotional responses in response to stress stimuli [17]. Our experiments have provided evidence supporting this hypothesis. By integrating conditional knockdown techniques with behavioral tests, this study confirms that dmPFC MORs are central to ARS-induced anxiety-like behavior. In MOR cKD mice, ARS-induced decrease in central exploration during OFT, as well as the time spent and entries in the open arms during EPM, were reversed. These results emphasize dmPFC MORs for inducing and sustaining elevated anxiety-like reactions under acute stress. Notably, dmPFC MOR systems react strongly to acute stress [17]. During acute stress, MORs in the CNS can quickly enhance the release of EOS, such as beta-endorphin and endomorphin [32]. Thus, acute stress-induced analgesia (ASIA) and conditional fear memory are two stress-related mechanisms that have been implicated in this activation [33, 34]. This study builds on existing evidence by elucidating the role of dmPFC MORs in ARS-induced anxiety-like behavior. Together, these findings expand on the knowledge of MOR-mediated control in the dmPFC and could theoretically explain physiological as well as pathological behavioral processes. The stress response is an instinctive reaction that promotes rapid escape from noxious stimuli. It triggers protective effects such as analgesia and heightened vigilance, while also inducing transient anxiety-like emotional disturbances [35–37]. This suggests that stress both defends and, to some extent, drives avoidance of the stressor. Our findings highlight the involvement of dmPFC MOR in this process, providing a theoretical framework for better understanding the complex regulatory logic of adaptive responses to environmental challenges.

Interestingly, dmPFC MORs are expressed in both glutamatergic and GABAergic neurons in the CNS [17]. MORs frequently confer opposing physiological effects on these two subtypes [22]. Thus, the present study investigated how MOR_GABA and MOR_Glut contributed diversely to ARS-induced anxiety-like behavior. MOR_Glut did not regulate these processes, whereas MOR_GABA did. In behavioral tests, both MOR_Glut WT mice and MOR_Glut cKD mice showed consistently significant anxiety-like reactions after ARS exposure. In contrast, MOR_GABA cKD mice did not demonstrate strong anxiety-like reactions observed in MOR_GABA WT mice under ARS. These results support the predominant role of dmPFC MOR_GABA in ARS-induced anxiety-like behavior. This finding is corroborated by our previous and present morphological results, which showed abundant MOR expression in dmPFC GABAergic interneurons [17]. In contrast, only 30% of dmPFC glutamatergic neurons expressed MORs [17]. MOR_GABA likely serves as the primary target of dmPFC MOR activation, wherein MOR engagement suppresses GABAergic interneuron activity, consistent with our earlier findings [17]. Thus, this disinhibition can subsequently activate excitatory pyramidal neurons originating from the dmPFC [17, 38]. The function of dmPFC MORs in ARS-induced anxiety-like reactions appears to depend on this mechanism, which was further supported by c-fos staining, and whole-cell patch clamp observations. These results show promise for clinical translation. Specifically targeting MOR_GABA within the dmPFC microcircuit, as opposed to systemic opioid receptors, could provide a more refined therapeutic approach for anxiety with a reduced side effects. In addition to emotional regulation, the dmPFC is involved in numerous higher-order cortical processes, such as pain modulation, decision making and language processing. Previous work has shown that targeted knockdown of MOR in cortical vGlut1-positive neurons produces antinociceptive-like responses in mice under moderate stress conditions [24]. Thus, MOR_Glut may contribute to these sensory or cognitive functions, despite its limited role in ARS-induced anxiety-like reactions.

The present study further demonstrated the contribution of dmPFC-BLA projections in ARS-induced anxiety-like behavior and emphasized how dmPFC MOR_GABA regulates these projections. The activity of dmPFC-BLA projections is strongly correlated with emotional disorder responses. Optogenetic activation of dmPFC-BLA projections specifically induce anxiety-like behaviors in the OFT. In contrast, inhibition of these projections markedly attenuate anxiety in rodents [7]. By using an inhibitory chemogenetic approach, suppressing dmPFC-BLA projection activity significantly reversed ARS-induced anxiety-like behavior. This finding is consistent with other research showing that the dmPFC-BLA pathway is a vital top-down modulator of behavior resembling anxiety and/or depression [7]. Notably, increased anxiety after exposure to several stressors has been associated with enhanced activity of these projections [7]. Furthermore, optogenetic activation of dmPFC-BLA projections intensifies anxiety-like reactions in both OFT and forced swim tests [7]. Our findings are in line with previous research, emphasizing the implications of dmPFC-BLA projections in anxiety-related behavior. Specifically, increased activity of dmPFC-BLA projections was strongly correlated with greater anxiety-like reactions during ARS. Moreover, dmPFC MOR_GABA modulates the activity of dmPFC-BLA projections under ARS, as evidenced by the use of c-fos staining. MOR_GABA WT mice demonstrated a higher frequency of spontaneous calcium signals after ARS, whereas MOR_GABA cKD mice showed a reverse effect. Likewise, MOR_GABA cKD mice did not exhibit the increased c-fos expression in dmPFC-BLA projections under ARS, which was observed in MOR_GABA WT mice. These results offer a molecular rationale for the increased dmPFC-BLA projection activity observed during acute stress and are in line with this MOR_GABA-controlled “disinhibition” hypothesis [39, 40].

The present study further looked into the electrophysiological mechanism by which dmPFC MOR_GABA controls dmPFC-BLA projections using whole-cell patch clamp readings. MOR_GABA lowered the GABAergic inputs to dmPFC-BLA projections, as indicated by decreased mIPSC frequencies in MOR_GABA WT mice but unaltered frequencies in MOR_GABA cKD mice after ARS, compared with US controls. Interestingly, MORs are common inhibitory GPCRs that effectively stop the inflow of sodium and calcium into neurons [41]. Consequently, after ARS, a significant decrease in GABAergic inputs to dmPFC-BLA projections was observed. In summary, the present study shows that dmPFC MOR_GABA is central to anxiety-like behavior caused by ARS exposure by integrating c-fos staining, and whole-cell patch clamp readings. Specifically, stress activated dmPFC MOR_GABA, which in turn reduced GABAergic inputs to dmPFC-BLA projections. This increased the neural activity of these projections and induced anxiety-like responses during behavioral tests.

Despite its strengths, this study has several limitations. First, notably, the findings are currently limited to male mice. This is because, in comparison to male mice, female mice exhibit non-negligible differences in the MOR-dependent modulation of emotional responses [42, 43]. Therefore, to obtain a deeper awareness of the modulatory role of dmPFC MOR_GABA in anxiety-like responses after ARS, future research involving female mice is needed. Second, this study only explored the modulation of MOR_GABA in dmPFC-BLA projections. However, the dmPFC also sends pyramidal projections to other areas of the brain, including the ventral tegmental area, laterodorsal tegmental nucleus, and median raphe nuclei, which are involved in emotional control [44, 45]. More research is necessary to determine how these projections contribute to anxiety-like behavior during ARS and how dmPFC MOR_GABA regulates them.

Conclusions

In conclusion, our present work demonstrated the physiological function of dmPFC MOR_GABA in controlling ARS induced anxiety-like responses by modulating of the activity of dmPFC-BLA projections. To decrease inhibitory inputs to dmPFC-BLA projections, the ARS efficiently enhanced the dmPFC MOR_GABA. Eventually, these projections became more active. Additionally, the active dmPFC-BLA pathway triggered by this disinhibition resulted in the expression of anxiety-like behavior. Through presynaptic disinhibition of dmPFC-BLA projections, our research demonstrated a projection-specific mechanism that correlates dmPFC MOR_GABA to ARS-induced anxiety reactions. By identifying a specific biochemical pathway by which dmPFC MORs influence anxiety-related behavior in acute stress, these results will provide a possible conceptual basis for diagnostic and therapeutic approaches targeted at anxiety disorders.

Author contributions

Yinan Du and Zhiqiang Liu designed experiments, conceived the project and prepared the manuscript. Yinan Du performed morphological tests. Aozhuo Zhang, Fenghua Xu, Junmin Zhang and Yukui Zhao performed behavioral tests and whole cell patch clamp experiments. Shuyi Liu, Yuan Chang, Yiwei Liu, Yingchun Yang, Yingying Li, Hongyu Qin, Chunling Wei, Qiaohua Zheng, Yanning Qiao, Yihui Liu, Wei Ren, Jing Han and Fei Gao helped to write and refine the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (grants 91949105, 82071516), Natural Science Basic Research Plan in Shaanxi Province of China (grants 2024JC-YBQN-0902), Excellent Graduate Training Program of Shaanxi Normal University (LHRCYB223009), and the Scientific Research Foundation of Xi’an Medical University (grants XYFYQDJF-2025-02).

Data availability

No datasets were generated or analysed during the current study.

Declarations

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. This manuscript is approved by all authors for publication.

Competing interests

The authors declare no competing interests.