A GABAergic pathway from dorsal raphe nucleus to paraventricular thalamic nucleus modulates incision-related pain behaviour in mice
Huijie Zhang, Lei Li, Bo Sun, Yinxiu Gao, Jingjing Zhang, Ce Bian, Yibo Wang, Man Li, Songxue Su, Weidong Zang, Jing Cao
TL;DR
A specific brain pathway involving GABAergic neurons from the dorsal raphe nucleus to the paraventricular thalamic nucleus helps reduce post-surgery pain in mice.
Contribution
The study identifies a novel GABAergic neural circuit that modulates incision-related pain behavior through GABAA receptors in mice.
Findings
Activation of the DRNGABA−PVT pathway alleviates incision pain in mice.
GABAA receptor antagonists in the PVT block the analgesic effects of this pathway.
The pathway is enhanced by external nociceptive stimuli and modulates neuronal excitability in the PVT.
Abstract
Incision pain is a prevalent condition in clinical practice, affecting approximately 50% of patients and significantly diminishing their quality of life. However, the central mechanisms underlying incision pain remain unclear. Here, we established a paw incision model that increased neuronal excitability in the paraventricular thalamic nucleus (PVT). Multiple tracing methods revealed an inhibitory ascending neural pathway from the dorsal raphe nucleus (DRN) to the PVT, with external nociceptive stimuli enhancing the activity of this pathway. Inhibition of the DRNGABA−PVT pathway induced nociceptive sensitivity in normal mice, while activation of this pathway alleviated incision pain. Notably, antagonists targeting GABAA receptors—not GABAB receptors—administered into the PVT blocked DRNGABA−PVT activation and produced significant analgesic effects on incision pain. Collectively, these…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7| Type (species) or virus strains | Designation | Source | Identifiers | Additional information |
|---|---|---|---|---|
|
|
| Jackson Laboratory | Stock No. 005359 | - |
|
|
| Jackson Laboratory | Stock No. 028867 | - |
| Genetic reagent | AAV2/5-hEF1a-DIO-H2B-eGFP-T2A-TVA-WPRE-pA | Taitool, Shanghai, China | Cat# S0320-5-H5 | 2× 1013 v.g./mL |
| Genetic reagent | AAV2/5-hEF1a DIO-RVG-WPRE-pA | Taitool, Shanghai, China | Cat# S0320-5-H5 | 1× 1013 v.g./mL |
| Genetic reagent | EnvA-ΔG-DsRed | Taitool, Shanghai, China | Cat# S0320-5-H5 | 1× 109 v.g./mL |
| Genetic reagent | AAV2/9-hSyn-DIO-mGFP-T2A-Synaptophysin-mRuby-WPRE-hGH | Brain VTA, Wuhan, China | Cat# PT-1244 | 5× 1012 v.g./mL |
| Genetic reagent | AAV2/2Retro-hSyn-Cre-WPREs | Brain VTA, Wuhan, China | Cat# PT-0042 | 2× 1012 v.g./mL |
| Genetic reagent | AAV2/9-EF1a-DIO-hM3Dq-mCherry | Brain VTA, Wuhan, China | Cat# PT-0988 | 2× 1012 v.g./mL |
| Genetic reagent | AAV2/2Retro-hSyn-Cre-WPREs | Brain VTA, Wuhan, China | Cat# PT-0042 | 2× 1012 v.g./mL |
| Genetic reagent | AAV2/9-EF1a-DIO-hM4Di-EGFP | Brain VTA, Wuhan, China | Cat# PT-0815 | 2× 1012 v.g./mL |
| Genetic reagent | AAV2/2Retro-DIO-flp | Brain Case, Shenzhen, China | Cat# BC-0718 | 5× 1012 v.g./mL |
| Genetic reagent | AAV2/9-hEF1a-fDIO-hM3Dq-EGFP-EPRE | Brain Case, Shenzhen, China | Cat# S0748-9-H2O | 2× 1012 v.g./mL |
| Genetic reagent | pAAV2/8-EF1a-fDIO-hM4Di-mCherry | OBIO, Shanghai, China | Cat# OP1028 | 2× 1012 v.g./mL |
| Genetic reagent | pAAV-EF1a-DIO-hChR2 (H134R)-mCherry | OBIO, Shanghai, China | Cat# OPO526 | 5× 1012 v.g./mL |
| Genetic reagent | pAAV2/9-EF1a-fDIO-ChR2-EGFP | OBIO, Shanghai, China | Cat# E2256 | 4× 1012 v.g./mL |
| Genetic reagent | rAAV-EF1a-FLEX-GCaMP7s-WPRE | OBIO, Shanghai, China | Cat# RA0721 | 2× 1012 v.g./mL |
- —National Natural Science Foundation of China10.13039/501100001809
- —Henan Provincial Science and Technology Research Project10.13039/501100017700
- —Henan Province Medical Science and Technology Public Relations Plan Province Department joint construction project
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsPain Mechanisms and Treatments · Anesthesia and Pain Management · Pediatric Pain Management Techniques
Introduction
Pain is a major medical concern leading to insomnia, anxiety, depression and other symptoms,^1-3^ yet current treatment modalities often fail to achieve satisfactory outcome. Studies suggest that brain pathways are closely associated with the development of chronic pain. Nevertheless, the specific neural pathways mediating chronic pain remain unknown. Therefore, clarifying the neural pathway involved in chronic pain might offer potential therapeutic targets.
Recently, the paraventricular thalamic nucleus (PVT), mainly composed of glutamatergic neurons, has attracted increasing attention for its function in pain sensation.^4-6^ For instance, Zhang et al.^7^ demonstrated that inhibition of the PVT–nucleus accumbens (NAc) pathway alleviates inflammatory pain in mice. Liang et al.^8^ reported that the excitatory neurons in PVT could activate GABAergic neurons of central nucleus of the amygdala (CeA) and transmit signals to the ventrolateral periaqueductal grey (vlPAG), thereby contributing to chronic pain. Moreover, PVT also receives a diverse array of excitatory inputs from various brain regions such as the medial prefrontal cortex (mPFC),^9^ locus coeruleus (LC),^10^ parabrachial nucleus (PBN),^11^ vlPAG and bed nucleus of the stria terminalis (BNST).^12,13^ Although the dysfunction of PVT neurons may arise from an imbalance between presynaptic excitation and inhibition transmission, research on inhibitory inputs related to pain processing within the PVT remains limited.
The dorsal raphe nucleus (DRN) is recognized as a key structure involved in pain modulation, with extensive evidence supporting its functional significance.^14-16^ Located ventrally to the periaqueductal grey (PAG), the DRN is implicated in endogenous pain-inhibitory pathways that regulate the transmission of nociceptive signals.^7,17,18^ Retrograde tracing studies have identified that the suprachiasmatic nucleus (SCN), the zona incerta (ZI) and the DRN as sources of inhibitory input to the PVT.^19^ Recent studies have shown that GABAergic neurons within the DRN play a crucial role in the regulation of wakefulness during stress by modulating the excitability of neurons in the PVT.^7^ Nevertheless, being a subordinate cell type within the DRN, GABA has been reported to be antianalgesic in the DRN.^20^ However, the involvement of DRN^GABA^−PVT pathways in pain processing has yet to be clarified.
In this study, we identified the activation of PVT^CaMKIIα^ neurons in incision pain conditions, and acknowledged the presence of DRN^GABA^−PVT^CaMKIIα^ neural pathway involved in pain processing. Fiber-photometry imaging showed that DRN neurons projecting to the PVT were activated by noxious stimuli. By employing chemogenetics and optogenetics strategies, we elucidated that inhibition of the DRN^GABA^–PVT pathway heightened pain sensitivity in normal mice, whereas activation of this pathway alleviated incision pain. Furthermore, pharmacological approaches demonstrated that intra−PVT administration of a GABA_A_ receptor antagonist reversed the analgesic effects induced by DRN activation. These findings suggest that the DRN exerts an analgesic influence by modulating the excitability of PVT neurons through inhibitory mechanisms.
Materials and methods
Animals
Male mice of the C57BL/6, GAD2-Cre and CaMKIIα-Cre strains aged 8∼10 weeks (the weight of mice = 20∼25 g) were utilized in this study. During all experiments, these mice were housed, 3∼5 per cage in a colony, in a stable environment (23∼25°C) with ad libitum access to standard lab mouse pellet food and water under a 12 h light/12 h dark cycle (lights on from 08:00 to 20:00). All experimental procedures were in accordance with the National Institutes of Health guidelines and were approved by the Zhengzhou University Animal Care and Use Committee.
Models of incision pain
We randomly assigned mice to control and experimental groups using a random number method. All eligible mice were numbered (1, 2, 3… N), sorted by random numbers generated in Excel, and then allocated sequentially: the first 50% to the control group and the remaining 50% to the experimental group. For more groups (e.g. three groups), the sorted list was divided proportionally (first, middle and last thirds). The experimental mice were anaesthetized with 2% isoflurane. Following disinfection using a 10% povidone−iodine solution, a 5-mm lengthwise incision was made on the left hind paw with a blade to deactivate and separate the tendo.^21^ After mild pressure was applied to stop the bleeding, the wound was sutured with nylon thread and sterilized. Control mice were anaesthetized and prepared identically to those in the incision group but did not receive the surgical incision. After the operation, the mice were allowed to be fully awake and return to the cage.
Stereotactic surgery and microinjection
The mice were anaesthetized with isoflurane and then secured in a stereotactic apparatus with their skull exposed. The virus was then injected into the PVT or DRN at a rate of 15–20 nL/min (coordinates: PVT: anterior−posterior (AP) = −1.55 mm, medio-lateral (ML) = 0 mm, dorsoventral (DV) = −3.10 mm; DRN: AP = −4.60 mm, ML = 0 mm, DV = −3.00 mm). A waiting period of an additional 10 min was implemented to prevent diffusion and backflow. Behavioural assessments were conducted three weeks post-injection. For retrograde monosynaptic tracing, the helper viruses AAV2/5-hEF1a-DIO-H2B-eGFP-T2A-TVA-WPRE-pA (80 nL) and AAV2/5-hEF1a DIO-RVG-WPRE-pA (80 nL) were mixed into the PVT of CaMKIIα-Cre mice. After three weeks, EnvA-ΔG-DsRed (100 nL) was injected into the same site. For anterograde tracing, AAV2/9-hSyn-DIO-mGFP-T2A-Synaptophysin-mRuby-WPRE-hGH (100 nL) was injected into DRN of the GAD2-Cre mice. Synaptophysin serves as a synaptic marker protein located on synaptic vesicles and was utilized to delineate the location of synapses.^22^ mGFP was employed to label both the somas and axons of DRN GABAergic neurons.
For chemogenetic manipulation, the virus AAV2/2Retro-DIO-flp (200 nL) was injected into the PVT, and AAV2/9-hEF1a-fDIO-hM3Dq-EGFP-EPRE (200 nL) or pAAV2/8-EF1a-fDIO-hM4Di-mCherry (200 nL) or its control (vector) were injected into DRN of GAD2-Cre mice. Before local drug infusion, the virus AAV2/2Retro-hSyn-Cre-WPREs (200 nL) was injected into the PVT and the AAV2/9-EF1a-DIO-hM3Dq-mCherry (200 nL) or AAV2/9-EF1a-DIO-hM4Di-EGFP (200 nL) was injected into the DRN.
For optogenetic manipulation, the virus pAAV-EF1a-DIO-hChR2 (H134R)-mCherry (200 nL) or its control (vector) were injected into DRN of GAD2-Cre mice. After 2 weeks, optic fibres were implanted over the PVT. We also injected the virus AAV2/2Retro-DIO-flp into the PVT of GAD2-Cre mice, pAAV2/9-EF1a-fDIO-ChR2-EGFP (200 nL) into the DRN, and optic fibres were implanted over the DRN to specifically activate DRN^GABA^-PVT.
For fibre-photometry experiments, rAAV-EF1a-FLEX-GCaMP7s-WPRE (300 nL) and AAV2/2Retro-hSyn-Cre-WPREs (200 nL) viruses were injected into the DRN and PVT of the wild-type mice, respectively. The optic fibres were implanted over the DRN. (Table 1)
Optogenetic manipulation
The optoelectrode was implanted over the PVT nucleus (AP, −1.55 mm, ML, 0 mm, DV, −2.90 mm). Following a one-week recovery period, behaviour tests with optical stimulation were started. The fibres were connected to a laser generator using optic fibre sleeves and a 5-min pulse of blue light (473 nm, 15 ms, 20 Hz, 5 mW) was delivered. Following verification of the injection sites via immunofluorescence staining, animals with incorrect placements were excluded from the behavioural study. To maintain a consistent group size, additional mice were introduced to reach eight per group.
Fibre-photometry imaging
The fibre-photometry system (R810, RWD Life Science, Shenzhen, China) was employed to record fluorescence signals generated by 470 and 410 nm LED light sources. During periods of inactivity, the basal fluorescence of mice was recorded for 5 min. Then, the mice were subjected to stimulation using von Frey filaments (0.07 g, 1.0 g), tail pinch and acetone application. Fluorescence changes were quantified by calculating ΔF/F values; specifically, the area under the curve (AUC) was determined from 2 s prior to stimulus onset to 10 s following stimulus onset. ΔF/F was calculated as (470 nm signal—fitted 410 nm signal) / (fitted 410 nm signal), where the fitted 410 nm signal represents baseline fluorescence intensity. The ΔF/F signals obtained during a period of 15 s before stimulus application served as baseline measurements. The selected duration for establishing this baseline encompassed data collected from 2 s before to 10 s after each event.
Drug infusion
Intraperitoneal administration
Clozapine-N-oxide (CNO) was dissolved in sterile normal saline and administered intraperitoneally (i.p.) at a dose of 2 mg/kg.
Local drug infusion
For microinjection of drugs into the PVT, a cannula (4 mm, RWD) was implanted over the PVT with a catheter clamp and fixed to the skull with dental cement, and then the catheter cap was screwed onto the catheter. After 1 week of recovery, an internal stainless-steel injector attached to a 1 μL syringe (Hamilton) was inserted into the guide cannula for infusing selective GABA_A_R antagonist bicuculline (0.1 μg), selective GABA_B_R antagonist CGP55845 (1 µmol) or saline (500 nL) into the PVT 40 min before testing. The mice were tested for pain behaviours after infusion.
Pain behaviours
For 3 days prior to the experiment, the mice were placed on the metal frame of an organic transparent resin box and habituated for 30 min every day. Mechanical, cold and thermal sensitivity to pain were tested by using mechanical paw withdrawal frequency (PWF), acetone score and Hargreaves test. For the incision pain mice, all behavioural tests were measured on postoperative Days 1, 2 (following CNO administration) and 3.
Mechanical PWF
When the mice were quiet, their left posterior paw was stimulated with 0.07 g and 0.4 g von Frey and the retraction was observed.^23^ Any instances of walking, flicking or licking of the feet were recorded as positive reaction ‘Χ’, while the absence of such responses was noted as a negative reaction ‘O’. The percentage of positive reaction was calculated. Stimulation trials were repeated 10 times with a 10 s interval between each trial, and each stimulus lasted for 2 to 3 s.
Acetone score
Acetone (20 μL) was applied to the skin of the left hind paw and the withdrawal reactions were observed within 40 s.^24^ Responses were then scored as follows: 0 points for no response; 1 point for a quick retreat, flick or stomp; 2 points for prolonged withdrawal or repeated licking of the paw and 3 points for continuous licking and slapping of the paw. The experiment was conducted three times with a 5-min interval between each trial, and the final score was the average of the three trials.
Hargreaves assay
When the mice were quiet, radiant heat was delivered to the middle of the left hind paw through the glass plate. The beam automatically cut-off when the mouse lifted its foot, and the duration from the onset of the beam to hind paw lifting was recorded as thermal paw withdrawal latency (PWL). The basal PWL in naïve mice was adjusted to 10–12 s and the maximum cut-off was set at 20 s to prevent any tissue damage. Each experiment was conducted three times with a 10-min interval between trials, and the final score represented the average of these three trials.
Electrophysiology
Electrophysiological recordings were obtained as described previously.^8^ Three weeks after viral injections into the DRN and PVT, mice weighing 20–25 g were anaesthetized with isoflurane and euthanized for brain extraction. Coronal slices (400 μm) containing the PVT were prepared under cold conditions (2∼4°C) using a vibratome (Leica VT, Germany). Only one neuron per slice within the PVT was recorded. The slicing solution contained the following (in mM): 2.5 KCl, 1.2 NaH2PO4, 26NaHCO3, 220 sucrose, 6 MgSO4, 0.5 CaCl2 and 10 glucose (pH 7.4) bubbled with carbogen gas (95% O_2_ and 5% CO_2_). Then, the slices were transferred to artificial cerebrospinal fluid (aCSF) containing (in mM): 124 NaCl, 2.5 KCl, 2 MgSO₄, 1 NaH₂PO₄, 25 NaHCO₃, 2 CaCl₂ and 37 glucose (pH 7.4).
Subsequently, slices were transferred to a 0.5 mL recording chamber mounted on a fixed-stage upright microscope (BX51W1, Olympus, Tokyo, Japan). Patch pipettes were fabricated from borosilicate glass capillaries (outer diameter: 1.5 mm, inner diameter: 0.84 mm, with microfilaments; 1B150F-4, World Precision Instruments, Sarasota, FL, USA) using a P-97 puller (Sutter Instruments, Novato, CA, USA). The internal pipette solution contained (in mM): 130 K-gluconate, 5 NaCl, 15 KCl, 0.4 EGTA, 10 HEPES, 4 Mg-ATP and 0.2 Tris-GTP, yielding resistances of 3–5 MΩ. Whole-cell recordings were targeted to fluorescently labelled PVT neurons receiving GABAergic inputs from the DRN. Fluorescence imaging was performed using a TRITC filter set (U-HGLGPS, Olympus) coupled to a monochrome CCD camera (IR-1000E, DAGE-MTI, Michigan City, IN, USA). Signals were amplified with a Multiclamp 700B amplifier (Axon Instruments, Foster City, CA, USA), acquired and analyzed using pCLAMP software (v. 10.02, Axon Instruments). Data were low-pass filtered at 2.6 kHz and digitized at 10 kHz (Digidata 1322A, Axon Instruments). Prior to recording, neurons were stabilized with at least three stimulation trials. Neuronal excitability was assessed by measuring spontaneous excitatory postsynaptic currents (sEPSCs) and the minimal current required to evoke action potentials (threshold current). During baseline, 18 current injections were applied. The perfusate was then switched to aCSF containing 10 μM CNO and allowed to equilibrate for 3 min. Current-clamp recordings continued for 20 min under CNO using the same stimulation protocol.
Histology
The mice were deeply anaesthetized with isoflurane and perfused with 50 mL saline, followed by 50 mL 4% of paraformaldehyde. The brains were immersed into 4% paraformaldehyde at 4°C overnight. They were then placed in a 20% sucrose solution until fully saturated, followed by immersion in a 30% sucrose solution until saturation. Coronal sections with a thickness of 30 μm were prepared at −20°C using a freezing microtome (CM1950, Leica). Sections designated for immunofluorescence staining were rinsed three times (10 min each) in 0.01 M PBS. They underwent pre-incubation for 2 h in PBS containing 10% normal goat serum and 0.5% Triton X-100 at room temperature. The sections were treated with rabbit anti-Fos (1:1200, Abcam) or rabbit anti-CaMKIIα (1:500, Abcam) or rabbit anti-TH (1:500, Alomone) or rabbit anti-5-HT (1:600, ImmunoStar). The primary antibodies were incubated overnight on a shaker at 4°C. The sections were then treated with goat anti-rabbit antibody conjugated to Cy3 (1:200, Proteintech) or 488 (1:200, Proteintech) for 2 h in an incubator at 37°C. After the incubation, the slices were washed in PBS for three times (10 min each time) and dyed with DAPI. Images were captured using a DMI4000 fluorescence microscope equipped with a DFC365FX camera (Leica).
Analysis
All data were statistically analyzed using GraphPad Prism 9, and presented as mean ± SEM. Normality was assessed using the Shapiro-Wilk test prior to conducting analysis of variance (ANOVA). No statistical methods were used to predetermine sample sizes, but our sample sizes were similar to those reported in previous publications.^25,26^ Statistical significance was assessed using one-tailed or two-tailed Student's t-tests, and one-way or two-way ANOVA, as appropriate. A value of P < 0.05 was considered statistically significant.
Results
Mice with incision pain show increased CaMKIIα neuronal activity in the PVT
Baseline measurement of nociceptive behaviours was performed preoperatively, with follow-up assessments on postoperative Days 1, 3, 5 and 7. Quantitative analysis demonstrated significant mechanical hypersensitivity in the incision group, manifesting as increased PWF to 0.07 g and 0.4 g von Frey filament stimulation and enhanced acetone-evoked cold allodynia responses (Fig. 1A–C). Concomitant thermal hyperalgesia was evidenced by reduced paw withdrawal latency to radiant heat (Fig. 1D). These validated behavioural end-points confirmed successful incisional pain model establishment.
*Incision pain activates excitatory neurons in the PVT. A. Behavioural responses to 0.07 g mechanical stimuli at days 1, 3, 5 and 7 after incision pain model. B. Behavioural responses to 0.4 g mechanical stimuli at days 1, 3, 5 and 7 after incision pain model. C. The acetone score of mice at days 1, 3, 5 and 7 after incision pain model. D. The thermal PWL of the mice at days 1, 3, 5 and 7 after incision pain model. (A−D) data are presented as mean ± SEM (n = 8 mice per group). Incision group versus Control group, two-way ANOVA. 0.07 g PWF: F(4,70) = 17.07, P < 0.001; 0.4 g PWF: F(4,70) = 18.21, P < 0.001; acetone score: F(4,70) = 12.05, P < 0.001; PWL: F(4,70) = 24.64, P < 0.001. Each data point represents an individual mouse in response to a specific stimulus type. E. The representative images of c-Fos (left) and CaMKIIα (middle) immunostaining of the control and the incision groups at Day 1 after incision pain model, Scale bars, 100 µm. F. Quantification of c-Fos positive neurons in the PVT. Incision group versus Control group, two-tailed, unpaired t test: t(6) = 5.61, individual data point shows one slice, each slice from one mouse, n = 4. Data are presented as mean ± SEM. G. Percentage of c-Fos positive neurons co-localized with the excitatory neuronal marker CaMKIIα in the PVT. Incision group versus Control group, two-tailed, unpaired t test: t(6) = 4.91, n = 4 slices from four mice. ***P < 0.001, **P < 0.01, P < 0.05. Data are presented as mean ± SEM. BL, baseline; PWF, paw withdrawal frequency; PWL, thermal paw withdrawal latency.
Next, we explored the possible neural mechanism underlying incision pain. Given the role of PVT in pain sensation and the predominant neuronal population of glutamatergic neurons within the PVT,^27^ we assess changes in neuronal activity of glutamatergic neurons in the PVT among mice in the incision group (Fig. 1E). Immunofluorescence analysis revealed that mice subjected to paw incision showed a significant increase in c-Fos-positive neurons in the PVT compared to that of control groups (Fig. 1F). Strikingly, we identified that majority of these c-Fos-positive neurons were co-localized with CaMKIIα, a marker for excitatory neurons^28^ (Fig. 1G),suggesting that paw incision led to the enhanced neuronal activity of CaMKIIα in the PVT.
An inhibitory pathway from DRNGABA to PVTCaMKIIα
To identify inhibitory inputs to the PVT and elucidate endogenous analgesic pathways, we unilaterally injected the retrograde tracer AAV2/2-Retro-EF1α-mCherry into the PVT and assessed afferent projections (Supplementary Fig. 1A, B). Three weeks post-transfection, a time point considered as indicative of stable viral transduction and expression, we observed a substantial population of mCherry-expressing neurons in the DRN, approximately 69% of which exhibited GABAergic positivity, indicating that the DRN neurons projecting to the PVT are primarily GABAergic neurons. Furthermore, consistent with the previous findings, we noted that approximately 6.7% of the mCherry-labelled neurons co-localized with serotonin (5-hydroxytryptamine, 5-HT), while 2.4% co-localized with tyrosine hydroxylase (TH) (Supplementary Fig 1C–F).
To validate this projection, we employed transgenic CaMKIIα-Cre mice and administered the helper adenoviral vectors AAV-DIO-TVA-GFP and AAV-DIO-RG into the PVT. Following a 21-day interval, we introduced RV-EnvA-ΔG-DsRed viral vector at the same injection site as illustrated in Fig. 2A. As expected, a significant number of DsRed-positive neurons were detected within the DRN, symmetrically positioned along the median plane (Fig. 2B). Notably, approximately 73.9 ± 13.3% of these neurons demonstrated co-localization with GABAergic (Fig. 2C and D), indicating the innervation from DRN GABAergic neurons to PVT CaMKIIα neurons.
An inhibitory pathway from DRNGABA to PVTCaMKIIα. A. Schematic of the retrograde trans synaptic virus (RV) injection. B. Viral expression in the PVT and DRN of CaMKIIα-Cre mice. Scale bars, 100 µm (left and middle) and 50 µm (right). C. DsRed-labelled neurons of DRN were co-localized with GABA neurons. Scale bars, 50 µm. D. Percentage of DsRed positive neurons expressing GABA. n = 3 mice. E. Schematic of viral injection. F. Typical images showing the neurons within the DRN expressing mRuby and mGFP. Scale bars, 100 µm. G The termination of axons of DRNGABA neurons within the PVT, including anterior (PVA: AP = −0.83 mm), middle (PV: AP = −1.23 mm and AP = −1.55 mm) and posterior (PVP: AP = −1.79 mm). Scale bars, 100 µm. Aq, aqueduct; AP, anterior−posterior; RV, retrograde trans synaptic virus; PVA, anterior part of PVT; PVP, posterior part of PVT; PVT, paraventricular thalamic nucleus; DRN, dorsal raphe nucleus.
To further substantiate the inhibitory pathway from DRN to PVT, we elucidated the synaptic connections between these structures. AAV2/9-hSyn-DIO-mGFP-T2A-Synaptophysin-mRuby was injected into the DRN of GAD2-Cre mice, which labels both somatic cells and synapses (Fig. 2E and F). A considerable number of axonal terminals from DRN GABAergic neurons was identified within the PVT, suggesting that DRN sends inhibitory fibres to the PVT^CaMKIIα^ neurons (Fig. 2G). Furthermore, functional studies further confirmed the inhibitory projection from DRN to PVT. This was observed from the increased expression of c-Fos positive neurons within DRN but the decreased expression of c-Fos positive neurons within PVT among mice receiving AAV2/9-hSyn-DIO-hM3Dq-EGFP and CNO injection when compared to the control groups with incision pain (Supplementary Fig. 2).
GABAergic ablation within DRN leads to pain hypersensitivity in naïve mice
To address the specific functional contributions of DRN GABAergic neurons to pain processing, we injected AAV2/9-EF1α-DIO-DTA into the DRN of GAD2-Cre mice under normal conditions, thereby achieving targeted ablation of GABAergic neurons, as illustrated in (Fig. 3A and B). Notably, the ablation of these GABAergic neurons results in an augmented responsiveness to both mechanical and cold stimuli, as demonstrated by the marked decrease in mechanical PWF at both 0.07 g and 0.4 g, along with increased acetone sensitivity scores (Fig. 3C–E), without altering thermal nociceptive thresholds (Fig. 3F). This functional specificity is consistent with prior reports on the role of DRN GABAergic signalling in modulating distinct sensory modalities.^29^
*Ablation of the GABA neurons in DRN increases pain hypersensitivity in naïve mice. A. Schematic of viral injection. B. Representative images of GABA neurons in DRN immunofluorescence for control and DTA group. Scale bars, 50 µm. C–F. Behavioural effects of ablation of DRNGABA neurons on pain behaviours assessed by von Frey, acetone and thermal PWL test. Individual data point shows one mouse. DTA group versus Control group, unpaired t-test, 0.07 g PWF: t(14) = 2.263, P < 0.05; 0.4 g PWF: t(14) = 2.614, P < 0.05; acetone score: t(14) = 4.235, P < 0.001; PWL: t(14) = 1.797. ***P < 0.001, P < 0.05. n = 8 mice per group. Data are presented as Mean ± SEM. Aq, Aqueduct; DTA, diphtheria toxin A subunit; DRN, dorsal raphe nucleus; PWF, paw withdrawal frequency; PWL, thermal paw withdrawal latency.
DRNGABA neurons projecting to PVT were activated by noxious stimuli
We have revealed the inhibitory pathway from DRN to PVT and evaluated the effect of DRN GABAergic neurons in pain modulation, however, the contribution of DRN^GABA^-PVT signalling in response to noxious stimuli and pain processing remains unexamined. To address this, we investigated the activity dynamics of DRN GABAergic projections to the PVT during nociceptive stimulation using in vivo calcium imaging. AAV2/2-Retro-Cre was injected into the PVT, and rAAV-EF1α-FLEX-GCaMP7s was delivered to the DRN of GAD2-Cre mice (Fig. 4A–C), enabling GCaMP7s expression specifically in DRN neurons projecting to the PVT. Calcium signals were recorded in real time to monitor neuronal activity elicited by various somatosensory stimuli. Exposure to a noxious mechanical stimulus (1.0 g von Frey filament) evoked a significant increase in calcium transients, as reflected by elevated peak ΔF/F and area under the curve (AUC) values (Fig. 4H–K). Similar responses were observed following tail pinch (Fig. 4L–O) and acetone application (Fig. 4P–S). In contrast, innocuous stimulation with a 0.07 g von Frey filament failed to elicit significant calcium responses, with no changes detected in peak ΔF/F or AUC (Fig. 4D–G). These results demonstrate that DRN GABAergic projections to the PVT are selectively activated by noxious stimuli under physiological conditions, supporting their functional engagement in nociceptive processing.
*The DRN neurons projecting to the PVT responds to nociceptive stimuli. A. Schematic of viral injection. B. Schematic of the experimental timeline. C. A representative image of GCaMP7s expression within DRN (AP = −4.60 mm). Scale bars, 100 µm. D. Changes of calcium signal by 0.07 g von Frey stimuli. D. The heatmap illustrates the fluorescence activity of GCaMP6s, which corresponds to all instances of hind paw lifting in mice. The colour scale on the right indicates ΔF/F values. E: Mean fluorescent signal in response to hindpaw withdrawal in all the mice recorded and fluorescence changes were quantified by ΔF/F values. F−G: Peak of ΔF/F and AUC. Two-tailed unpaired t-test: peak ΔF/F: t(10) = 0.45; AUC: t(10) = 0.60. H. Changes of calcium signal by 1.0 g von Frey stimuli. H−K. Same conventions as D−G. Two-tailed unpaired t-test: peak ΔF/F: t(10) = 4.02; AUC: t(10) = 3.86. L. Changes of calcium signal by tail pinch stimuli. L−O. Same conventions as D−G. Two-tailed unpaired t-test: peak ΔF/F: t(10) = 2.73; AUC: t(10) = 2.46. P. Changes of calcium signal by acetone stimulation. P−S. Same conventions as D−G. Two-tailed unpaired t-test: peakΔF/F: t (10) = 5.28; AUC: t (10) = 5.31. *P < 0.05, **P < 0.01, **P < 0.001, n = 6 mice per group. Individual data point shows each response to stimuli. AUC: area under the curve. Data are presented as mean ± SEM. Aq, Aqueduct; AP, anterior-posterior; BL, baseline; WT, wide type; PVT, paraventricular thalamic nucleus; DRN, dorsal raphe nucleus.
Activation of the DRNGABA–PVT alleviated incision pain of mice
Next, we characterized the impact of DRN-mediated modulation on PVT neurons and their effects on pain responses. The AAV2/1-hSyn-Cre virus was administered into the DRN, while AAV2/9-hSyn-DIO-hM3Dq-mCherry, hM4Di or a control virus was into the PVT to selectively activate or inhibit PVT (Supplementary Fig. 3A). As expected, hM3Dq induced increased c-Fos staining while decreased c-Fos staining within PVT was detected among mice receiving hM4Di (Supplementary Fig. 3B, C). Behavioural evaluations disclosed that hM3Dq-mediated PVT activation engendered pain-like behaviour in response to mechanical and cold stimuli in naïve mice (Supplementary Fig 3D–F), while no significant modification occurred in sensitivity to thermal stimuli (Supplementary Fig. 3G). Furthermore, hM4Di- mediated PVT inhibition exerted no influence on behavioural phenotypes in naïve mice (Supplementary Fig 3D–G), conversely, it significantly mitigated pain-like behaviours including mechanical and cold allodynia, as well as thermal hyperalgesia among mice with incision pain (Supplementary Fig 3H–M).
To specifically assess the effect of DRN^GABA^–PVT pathway in pain regulation, we selectively activated this pathway through the utilization of optogenetic approaches and identified the effects of such activation on pain behaviour. We injected AAV2/2Retro-DIO-flp into the PVT and AAV-fDIO-ChR2-EGFP or control virus into the DRN of GAD2-Cre mice to selectively activate the DRN^GABA^–PVT projections (Fig. 5A and D). The detailed test procedures were presented in Fig. 5B and C. Compared to incision mice receiving the control virus, optogenetic activation of DRN^GABA^–PVT significantly reduced elevated mechanical PWF in response to both 0.07 g and 0.4 g stimuli, as well as diminished the increased acetone score induced by paw incision, thereby suggesting that activation of DRN^GABA^–PVT effectively alleviates incision pain (Fig. 5E–G). Likewise, activation of DRN^GABA^–PVT by another optogenetic activation strategy produced similar effects. By implanting optical fibres in the PVT, and injecting AAV2/9-EF1α-DIO-ChR2-mCherry or AAV2/9-EF1α-DIO-mCherry virus into the DRN of GAD2-Cre mice (Fig. 5H and I), we discovered that optogenetic activation of DRN^GABA^–PVT significantly effectively alleviates incision pain (Fig. 5J–L).
*Optogenetic activation of the DRNGABA−PVT alleviates incision pain in mice. A, H. Schematic of viral injection. B. The illustration of optogenetic activation. C. Optogenetic activation via a 473 nm laser (20 Hz, 5 mW, 15 ms). D. Representative images of fDIO-ChR2-EGFP. Scale bars, 100 µm. E−G. Behavioural effects of optogenetic activation of DRNGABA−PVT on pain behaviours assessed by von Frey and acetone test in incision pain mice. Two-way ANOVA, 0.07g PWF: F(2,42) = 8.06, P = 0.0011; 0.4 g PWF: F(2,42) = 9.125, P < 0.001; acetone score: F(2,42) = 6.80, P = 0.0028, n = 8 mice per group. I. Representative images of DIO-ChR2-mCherry. Scale bars, 100 µm. J−L. Same conventions as E−G. Two-way ANOVA: 0.07 g PWF: F(2,42) = 2.636, P = 0.0835; 0.4 g PWF: F(2,42) = 7.299, P < 0.001; acetone score: F(2,42) = 6.527, P = 0.0034. Individual data point shows one mouse. *P < 0.05, **P < 0.01, **P < 0.001, n = 8 mice per group. Data are presented as Mean ± SEM. Aq, Aqueduct; AP, anterior−posterior; ChR2, channelrhodopsin-2; PVT, paraventricular thalamic nucleus; DRN, dorsal raphe nucleus; PWF, paw withdrawal frequency.
Inhibition of the DRNGABA–PVT pathway pain-like behaviours in naïve mice
To better understand the regulatory effects of the DRN^GABA^–PVT pathway in pain modulation, we injected AAV2/2Retro-DIO-flp that was infused into the PVT while either AAV-fDIO-hM3Dq or hM4Di was delivered to the DRN of GAD2-Cre mice, to selective activate or inhibit DRN^GABA^–PVT under normal conditions. Anatomical analysis revealed that the labelled neurons were predominantly located in the DRN (Fig. 6A and B). The functional efficacy of hM3Dq and hM4Di expression was confirmed by increased or decreased spike frequencies before and after CNO administration, respectively (Supplementary Fig. 4). Behavioural results demonstrated that inhibition of the DRN^GABA^–PVT pathway enhanced pain sensitivity to mechanical and cold stimuli in mice (Fig. 6C–E). The thermal PWL remained unchanged (Fig. 6F). Activating the DRN^GABA^–PVT pathway did not modify the baseline responses to mechanical heat and acetone stimuli (Fig. 6C–F; two-way ANOVA: same statistics as above). Then, we tested pain behaviours after surgery (Fig. 6G and H) and showed that the mechanical and cold hypersensitivity reduced as well as the heat sensitivity alleviated against incision-induced pain (Fig. 6I–L).
*Chemogenetic activation of the DRNGABA−PVT alleviates incision pain in mice. A, G. Schematic of viral injection. B. Representative images of hM3Dq, and hM4Di. Scale bars, 50 µm. C−F. The 0.07 g, 0.4 g mechanical PWF, acetone score and thermal PWL. Two-way ANOVA: 0.07 g PWF: F(2,42) = 12.71, P < 0.001; 0.4 g PWF: F(2,42) = 4.35, P = 0.0192; acetone score: F(2,42) = 8.71, P < 0.001; F(2,42) = 4.47, P = 0.0174. N = 8 mice per group. H. Procedures for the pain behavioural tests. I-L. The 0.07 g, 0.4 g mechanical PWF, acetone score and thermal PWL. Two-way ANOVA, 0.07 g PWF: F(2,42) = 8.469, P = 0.007; 0.4 g PWF: F(2,42) = 26.58, P < 0.001; acetone score: F(2,42) = 28.12, P < 0.001; PWL: F(2,42) = 8.02, P = 0.0018. N = 8 mice per group. Individual data point shows one mouse. *P < 0.05, **P < 0.01, **P < 0.001. Data are presented as mean ± SEM. Each data point represents an individual mouse in response to a specific stimulus type. Aq, aqueduct; BL, baseline; CNO, Clozapine-N-oxide; PVT, paraventricular thalamic nucleus; DRN, dorsal raphe nucleus; PWF, paw withdrawal frequency; PWL, thermal paw withdrawal latency.
The DRN exerts analgesic effects through GABAA receptors in the PVT
Although the analgesic effects of the DRN^GABA^–PVT projection have been documented, the underlying molecular mechanisms remain poorly understood. Given the potential role of GABA receptors in pain regulation^6,30^ and their status as downstream targets innervated by GABAergic projections from the DRN, we investigated whether the DRN mediates its analgesic effects through GABA receptors in the PVT. To achieve this, we combined pharmacological approaches with chemogenetic manipulation to explore whether the inhibition of GABA receptors could hinder the analgesic effects of chemogenetically activated DRN^GABA^–PVT. Specifically, we respectively employed antagonists of GABA_A_ and GABA_B_ receptors, and implanted catheters into the PVT for drug administration, where bicuculline (0.1 μg) is targeted at the GABA_A_ receptors and CGP55845 is targeted at the GABA_B_ receptors (Fig. 7A–C and Supplementary Fig 5A–C). The c-Fos immunofluorescence staining and cannula placement within the PVT indicated that the specific antagonists of the GABA_A_ receptors (Fig. 7D), and GABA_B_ receptors (Supplementary Fig. 5D) were reliably injected into the PVT. As anticipated, activation of the DRN–PVT pathway demonstrated significant analgesic effects during incision pain; however, these effects were markedly reduced by bicuculline treatment, suggesting that the analgesic effects of DRN–PVT pathway may be mediated by GABA_A_ receptors (Fig. 7E-H and Supplementary Fig 5E–H). Conversely, the administration of CGP55845 was ineffective in preventing the analgesic effects elicited by DRN–PVT activation. Importantly, bicuculline alone did not alter basal nociceptive thresholds in naïve animals, confirming pathway-specific modulation under pathological states.
*Suppression of GABAA receptors in the PVT reverses the analgesic effect of the DRN−PVT projections. A. Schematic of viral injection. B. Timeline of drugs injection. C. Representative images of hM3Dq. Scale bars, 100 µm. D. c-Fos immunofluorescence staining and cannula placement in the PVT. Scale bars, 250 µm. E-H. The 0.07 g, 0.4 g mechanical PWF, acetone score and thermal PWL following PVT injections of saline or bicuculline under normal and incision pain conditions. Individual data point shows one mouse. Two-way ANOVA, 0.07 g PWF: F(1,30) = 38.67, P < 0.001; 0.4 g PWF: F(1,30) = 22.27, P < 0.001; acetone score: F(1,30) = 41.44, P < 0.001; PWL: F(1,30) = 79.86, P < 0.001). CNO group versus CNO + Bicuculline group; n = 6 mice per group. *P < 0.05, **P < 0.01, **P < 0.001. Data are presented as mean ± SEM. Each data point represents an individual mouse in response to a specific stimulus type. Aq, aqueduct; AP, anterior−posterior; BL, baseline; Can., cannula; CNO, Clozapine-N-oxide; WT, wide type; PVT, paraventricular thalamic nucleus; DRN, dorsal raphe nucleus; PWF, paw withdrawal frequency; PWL, thermal paw withdrawal latency.
Discussion
Several key findings were demonstrated in the current study: (1) Incision pain resulted in the activation of PVT CaMKIIα neurons, which receive GABAergic projections from the DRN; (2) the ablation of DRN GABAergic neurons produced pain-like behaviours in naïve mice; (3) DRN^GABA^–PVT modulated paw incision-induced pain related behaviours and responded to noxious stimuli under normal conditions; (4) DRN^GABA^–PVT regulated pain-related behaviours by acting on GABA_A_ receptors in PVT. Collectively, these findings establish pivotal roles of DRN^GABA^–PVT in mediating pain-related behaviours.
The PVT, primarily constituted by glutamatergic neurons, assumes a significant role in regulating sleep,^31^ and feeding,^32^ and is closely related to pain processing.^7,8^ Our findings corroborate previous studies by demonstrating increased activation of PVT glutamatergic neurons during incisional pain (Fig. 1), and inhibiting PVT neurons produced effective pain relief (Supplementary Fig. 3). While extensive research has characterized excitatory inputs to the PVT from various brain regions.^4,6,33,34^ The functional significance of its inhibitory inputs in pain modulation remains poorly understood. Although inhibitory projections from the ZI and DRN to the PVT have been anatomically described, only the ZI−PVT pathway has been functionally linked to neuropathic pain regulation.^35^ Recent evidence implicates the DRN^GABA^–PVT plays a crucial role in the regulation of wakefulness during stress,^36^ yet its involvement in nociception remained unexplored.
The cellular heterogeneity of the DRN underlies its complex neuromodulatory functions.^37-39^ While serotoninergic (5-HT) mechanisms in DRN-mediated pain signalling are well established,^40,41^ the specific contributions of GABAergic populations have remained elusive. Here, we demonstrate that ablating DRN GABAergic neurons enhances pain sensitivity in naïve mice, challenging the prevailing view of their primary role as local inhibitors of 5-HT neurons (Fig. 3).^42,43^ Furthermore, calcium imaging studies further revealed that DRN^GABA^–PVT projection neurons exhibit stimulus-specific activation patterns during nociception (Fig. 4). These collective findings identify a novel GABAergic pain modulatory pathway from the DRN to PVT, whose upstream regulatory mechanisms represent an important direction for future research.
Subsequent investigation employing optogenetic and chemogenetic approaches revealed that selective activation of the DRN^GABA^–PVT pathway significantly attenuated incision-induced mechanical allodynia, cold hypersensitivity and thermal hyperalgesia, providing novel insights into endogenous pain control mechanisms. Paradoxically, pathway inhibition predominantly exacerbated mechanical and cold nociception without significantly altering thermal pain thresholds. This modality-specific modulation may originate from distinct neurobiological substrates. Firstly, it may be related to the intrinsic properties of PVT neurons. Noxious thermal stimuli are transient in nature, while PVT glutamatergic neurons are predominantly low-threshold burst-spiking neurons. This characteristic may make them more suited to encoding sustained mechanical or cold stimuli rather than rapid thermal stimuli, leading to less efficient encoding of acute heat pain.^44,45^ Secondly, functional segregation may emerge from target-specific pathway engagement, as PVT outputs to limbic regions and cortex preferentially process the affective salience of mechanical/cold stimuli, while thermosensory discrimination involves distinct thalamocortical projections to primary somatosensory regions. For instance, the PVT–NAc^46,47^or PVT–CeA^48^ pathway, which regulates the affective component of pain, may exhibit stronger encoding for mechanical and cold pain compared to thermal pain.
Bicuculline acts as a selective negative modulator of GABA_A_ receptors with anticonvulsant properties, primarily utilized in epilepsy research. Recent studies have shown that bicuculline can attenuate pain responses in formalin test.^49^ In this study, local administration of the GABA_A_ antagonist bicuculline selectively abolished the analgesia induced by DRN–PVT activation. We demonstrated that the pain-regulatory function of the DRN^GABA^–PVT pathway is specifically mediated by GABA_A_ receptors in the PVT. This points to a central role of GABA_A_ receptors, further validation using cell-specific knockout models and detailed receptor localization studies is warranted.
This study provides initial evidence for the role of the DRN–PVT pathway in modulating incision pain. However, we acknowledge several limitations. For example, due to the well-documented influence of gonadal hormone fluctuations on nociceptive sensitivity and central pain processing,^50^ this study was conducted exclusively in male C57BL/6 mice to minimize physiological variability. Technically, while behavioural and pharmacological approaches effectively demonstrated the functional consequences of manipulating this pathway, the absence of in vivo electrophysiological recordings means we lack direct, real-time insight into the firing patterns and population activity of DRN–PVT neurons during pain processing. Furthermore, our model does not account for key clinical complexities, such as sleep deprivation—a common comorbidity in post-surgical pain. Given the PVT's established role in sleep−wake regulation and the known impact of sleep disruption on pain threshold and mood, its interaction with the DRN–PVT pathway likely represents a significant, unmodelled variable that limits the generalizability of our findings. To address these gaps, future work will employ a mouse model incorporating sleep deprivation to mimic the clinically relevant comorbid state of pain-sleep disturbances. This will allow us to determine whether the DRN–PVT pathway retains its efficacy in regulating the transition and maintenance of incisional pain and its analgesic effects under these more complex.
Supplementary Material
fcag046_Supplementary_Data
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Shetty A, Delanerolle G, Cavalini H, et al A systematic review and network meta-analysis of pharmaceutical interventions used to manage chronic pain. Sci Rep. 2024;14(1):1621.38238384 10.1038/s 41598-023-49761-3PMC 10796361 · doi ↗ · pubmed ↗
- 2Yang LN, Tang M, Nüssler AK, Liu L, Yang W. Regulation of PVT-Ce A pathway in Deoxynivalenol-induced Anorexia and aversive-like emotions. Adv Sci (Weinh). 2025;12(16):e 2417068.40019402 10.1002/advs.202417068 PMC 12021098 · doi ↗ · pubmed ↗
- 3Michaelides A, Zis P. Depression, anxiety and acute pain: Links and management challenges. Postgrad Med. 2019;131(7):438–444.31482756 10.1080/00325481.2019.1663705 · doi ↗ · pubmed ↗
- 4Petrovich GD . The function of paraventricular thalamic pathwayry in adaptive control of feeding behavior. Front Behav Neurosci. 2021;15:671096.33986649 10.3389/fnbeh.2021.671096 PMC 8110711 · doi ↗ · pubmed ↗
- 5Iglesias AG, Flagel SB. The paraventricular thalamus as a critical node of motivated behavior via the hypothalamic-thalamic-striatal pathway. Front Integr Neurosci. 2021;15:706713.34220458 10.3389/fnint.2021.706713 PMC 8250420 · doi ↗ · pubmed ↗
- 6Zhang FC, Wei YX, Weng RX, et al Paraventricular thalamus-insular cortex pathway mediates colorectal visceral pain induced by neonatal colonic inflammation in mice. CNS Neurosci Ther. 2024;30(4):e 14534.37994678 10.1111/cns.14534 PMC 11017444 · doi ↗ · pubmed ↗
- 7Cai P, Wang FD, Yao J, et al Regulation of wakefulness by GAB Aergic dorsal raphe nucleus-ventral tegmental area pathway. Sleep. 2022;45(12):zsac 235.36161495 10.1093/sleep/zsac 235 · doi ↗ · pubmed ↗
- 8Liang SH, Zhao WJ, Yin JB, et al A neural pathway from thalamic Paraventricular nucleus to central Amygdala for the facilitation of neuropathic pain. J Neurosci. 2020;40(41):7837–7854.32958568 10.1523/JNEUROSCI.2487-19.2020 PMC 7548696 · doi ↗ · pubmed ↗