Abstract
Chronic visceral pain is a major challenge for both patients and health providers. Although the central sensitization of the brain is thought to play an important role in the development of visceral pain, the detailed neural circuits remain largely unknown. Using a well-established chronic visceral hypersensitivity model induced by neonatal maternal deprivation (NMD) in male mice, we identified a distinct pathway whereby the claustrum (CL) glutamatergic neuron projecting to the anterior cingulate cortex (ACC) is critical for visceral pain but not for CFA-evoked inflammatory pain. By a combination of in vivo circuit-dissecting extracellular electrophysiological approaches and visceral pain related electromyographic (EMG) recordings, we demonstrated that optogenetic inhibition of CL glutamatergic activity suppressed the ACC neural activity and visceral hypersensitivity of NMD mice whereas selective activation of CL glutamatergic activity enhanced the ACC neural activity and evoked visceral pain of control mice. Further, optogenetic studies demonstrate a causal link between such neuronal activity and visceral pain behaviors. Chemogenetic activation or inhibition of ACC neural activities reversed the effects of optogenetic manipulation of CL neural activities on visceral pain responses. Importantly, molecular detection showed that NMD significantly enhances the expression of NMDA receptors and activated CaMKIIα in the ACC postsynaptic density (PSD) region. Together, our data establish a functional role for CL→ACC glutamatergic neurons in gating visceral pain, thus providing a potential treatment strategy for visceral pain.
SIGNIFICANCE STATEMENT Studies have shown that sensitization of anterior cingulate cortex (ACC) plays an important role in chronic pain. However, it is as yet unknown whether there is a specific brain region and a distinct neural circuit that helps the ACC to distinguish visceral and somatic pain. The present study demonstrates that claustrum (CL) glutamatergic neurons maybe responding to colorectal distention (CRD) rather than somatic stimulation and that a CL glutamatergic projection to ACC glutamatergic neuron regulates visceral pain in mice. Furthermore, excessive NMDA receptors and overactive CaMKIIα in the ACC postsynaptic density (PSD) region were observed in mice with chronic visceral pain. Together, these findings reveal a novel neural circuity underlying the central sensitization of chronic visceral pain.
Introduction
Chronic visceral pain affects up to 20% of the worldwide population and provides both a significant therapeutic challenge and a route to understanding mechanisms in the nervous system (Elsenbruch et al., 2015). As a functional gastrointestinal disorder characterized by chronic visceral pain and altered bowel movements, irritable bowel syndrome (IBS) is a serious affliction that negatively impacts the patient's quality of life (Bellini et al., 2014; Y. Xiao et al., 2016). However, the pathologic mechanism of chronic visceral pain is unclear, and there is a lack of effective therapeutics for its treatment. Therefore, improving our understanding of the pathogenesis of chronic visceral pain and developing effective treatments are crucial for the well-being of these patients. Many factors such as inflammation, changes in bowel function, and early life stressors are reported to induce IBS (Videlock et al., 2009; O'Mahony et al., 2017). Neonatal maternal deprivation (NMD) is a well-established animal model mimicking the effects of early life stress on the development of emotional and social behaviors (Du et al., 2019). Moreover, NMD induces chronic visceral hypersensitivity in adult rats and mice (Y Xiao et al., 2016). Therefore, this research aims to explore the neural circuit mechanisms of chronic visceral pain in a mouse model of chronic visceral hypersensitivity induced by NMD.
The anterior cingulate cortex (ACC) is a core brain region that processes the sensory and emotional components of chronic pain in rodents and humans (M. Zhou, 2008; Bliss et al., 2016; Vogt, 2019). Basic and clinical studies have revealed significant changes in the structure and function of the ACC in both patients and animal models under chronic pain conditions (Maclullich et al., 2006). Ablation of this region from cingulotomy, along with genetic or pharmacological inhibition, effectively diminishes pain and its affective components. The results of functional magnetic resonance imaging also revealed that the ACC was activated in IBS patients (Mayer et al., 2009; Tillisch et al., 2011; Mayer et al., 2015). Merging evidence indicates that the ACC processes both somatic and visceral pain information (Fan et al., 2009; F.L. Chen et al., 2012). However, it is as yet unknown whether there is a specific neuron type of ACC gating visceral pain and a distinct neural circuit which helps the ACC to distinguish the visceral from somatic pain information. Recently, it has been reported that there is a neural circuit relationship between the claustrum (CL) and ACC (Fillinger et al., 2017; Smith et al., 2020), but the specific neural types within this circuit were not identified. Although the CL, an enigmatic brain structure lying between the insular cortex and the striatum (Mathur, 2014), is reciprocally connected with almost all cortical areas, including the motor, somatosensory, and prefrontal cortices (Goll et al., 2015; White et al., 2017; Nikolenko et al., 2021), its role in chronic pain remains unknown. Since the CL projects to ACC region and ACC was closely related to chronic pain, we therefore speculate that CL regulates visceral pain by modulating ACC activities.
Therefore, the purpose of this study was to explore whether there is a specific neural circuit between CL and ACC which controls chronic visceral pain. Using an integrative approach of viral tracing, optogenetics, chemogenetics and electrophysiology, we identified a novel neural circuit that the CL to the ACC pathway activation mediates visceral hypersensitivity induced by NMD in mice. Mechanistic studies indicated that excessive NMDA receptor (NMDAR) and overactive CaMKIIα in the postsynaptic density (PSD) region fraction of ACC are likely responsible for the central sensitization of chronic visceral hypersensitivity in the NMD mice. Collectively, we provide evidence to confirm that the glutamatergic CL-ACC pathway is necessary and sufficient for manifesting visceral pain, thus providing a potential treatment strategy by modulating specific neural circuitry for chronic visceral pain.
Materials and Methods
Animals
Adult male and female mice (C57BL/6, VGlu2-ires-Cre) were used for breed. The VGlu2-Cre mice were purchased from Shanghai model organisms. After delivery, the dams and male pups were selected for induction of chronic visceral hypersensitivity. The adult male offspring were housed one to five per cage depending on the experimental need. They were free access to water and food (standard mouse chow) and maintained under a 12/12 h light/dark cycle at a stable temperature (23–25°C). Considering that estrogen synthesized in female mice may affect visceral pain (Mulak and Taché, 2010; Sanoja and Cervero, 2010), we excluded adult female mice from the study. Handling of the animals was approved by the Institutional Animal Care and Use Committee at Soochow University and was strictly in accordance with the guidelines of the International Association for the Study of Pain.
Measurement of chronic visceral pain
The visceral hypersensitivity was induced by NMD as described previously (Li et al., 2012; Miquel et al., 2016). Briefly, male pups in the model group were separated from their mother for 3 h every day for two weeks, from postnatal day 2nd to 15th. Pups in the control (CON) were not handled. Visceral pain was assessed by the colorectal distention (CRD) threshold as described previously (Hu et al., 2013; P.A. Zhang et al., 2017). Briefly, a flexible balloon (2 cm) was inserted into the colon and rectum via the anus after the mice were anesthetized with isoflurane (RWD). The balloon was made of a surgical glove attached to tygon tubing, which was taped to each mouse's tail. The mice were allowed to recover for 30 min in small, isolated cubicles before CRD was performed. The balloon was slowly inflated with a sphygmomanometer until the tested mouse showed a significant abdominal withdrawal reaction. To detect CRD-induced c-Fos expression, the balloon was inflated with a pressure of 60 mmHg for a 20-s stimulation period followed by a 3-min rest. Five CRD stimulations were applied to each mouse. After completion of the last CRD stimulation, the mouse was allowed to rest for 1 h before transcardially perfused with 0.9% saline. Visceral pain was also measured by electromyographic (EMG) recordings as described previously (Winston et al., 2007). In brief, under anesthesia with isoflurane, two electrodes were implanted in the external oblique muscle and externalized behind the head. The mice were allowed one week to recover from the surgery. The EMG signal was amplified, filtered at 300 Hz, and digitized using Acknowledge software (Biopac Systems, Inc). The area under the curve (AUC) for EMG activity during each 20 s of distention was calculated using an in-house computer program. The net value for each distension was calculated by subtracting the baseline value derived from the AUC for the 20-s predistention period. Each mouse underwent EMG recording twice. The mean AUC of the EMG calculated using the two repeated measurements from each mouse was used for each pressure in the following statistical analysis.
Measurement of hind paw withdrawal threshold
To explore whether CL is involved in somatic pain process, we used complete Freund's adjuvant (CFA)-induced arthritis as a somatic pain model. To induce somatic pain, CFA (25 µl) was injected into the left hindpaw under brief isoflurane anesthesia as described previously (Zhu et al., 2019). The CON mice were injected with same volume of normal saline (NS). The somatic pain was measured by hind paw withdrawal threshold. Each mouse was placed in a transparent plastic chamber and habituated to the test environment for 3 d before testing. Before testing, mice were allowed to acclimate to the testing environment for 30 min. We tested the withdrawal threshold of the planta when stimulating the hindpaw's surface with the Von Frey hairs. Pain thresholds were tested immediately after termination of 1 min light stimulation with optogenetic manipulations (Zhu et al., 2019). The paw withdrawal threshold in response to von Frey filament stimulation of the hind paw was determined as described previously (H.H. Zhang et al., 2015). All behavioral studies were performed under blinded conditions.
Immunofluorescence staining
Mice were then deeply anesthetized and transcardially perfused with 0.9% saline, followed by 4% paraformaldehyde (PFA) in PBS (pH 7.4). Brains were carefully removed and postfixed in 4% PFA for an additional 3–4 h. The brains were then transferred to 10%, 20%, or 30% sucrose dissolved in PBS until they sank to the bottom of the container. The brains were sliced into 30-µm coronal sections using a freezing microtome (Leica). The sections were stored in the refrigerator at −30°C until c-Fos staining. Sections were incubated in 0.3% Triton X-100 for 1 h and blocked with 7% donkey serum for 1 h at room temperature. Sections were incubated overnight with primary antibodies at 4°C. The primary antibodies included anti-CaMKIIα (1:200, mouse, Cell Signaling, 6G9), anti-glutamate (1:200, rabbit, Sigma, G6642), anti-c-Fos (1:200, rabbit, Cell Signaling, 9F6), anti-c-Fos (1:200, mouse, Santa Cruz, sc-271243), and anti-GFP (1:200, goat, Abcam, ab6662). The brain sections were incubated with one of these primary antibodies at 4°C for 24 h, followed by incubation with the corresponding fluorophore-conjugated secondary antibodies at room temperature for 1 h. The secondary antibodies included anti-rabbit Alexa Fluor 488 (1:500, Invitrogen, A21206) or anti-mouse Alexa Fluor 555 (1:100, Invitrogen, A31570). The fluorescence signals were visualized using a Zeiss AXIO SCOPE A1 microscope. For quantification c-Fos data, three sections around the ACC or CL region were counted manually using ImageJ software. The sections used were at the same coordinates for each group (Wang et al., 2020). In this part of the experiment, one brain slice was taken from one mouse.
Subcellular fractions
The extraction of the PSD region of ACC was performed as described previously (Zeng et al., 2018). In brief, the tissues of ACC were homogenized in buffer A (0.32 m sucrose, 1 mm MgCl2, 1 mm PMSF and a protease inhibitor cocktail). Homogenates were passed through a filter to remove cell debris and centrifuged at 500 × g for 5 min in a fixed angle rotor to yield P1 and S1 fractions. P1 fractions were washed in buffer B containing 10 mm KCl, 1.5 mm MgCl2, 10 mm Tris-HCl (pH 7.4) and centrifuged at 500 × g for 5 min. Pellets were dissolved in buffer C containing 20 mm HEPES (pH 7.9), 25% glycerol, 1.5 mm MgCl2, 1.4 m KCl, 0.2 mm EDTA, 0.2 mm PMSF, 0.5 mm DTT and incubated on a shaker at 4°C for 30 min. After centrifugation at 12,000 × g for 10 min, the supernatant of P1 was collected as nuclear proteins. The S1 fraction was centrifuged at 10,000 × g for 10 min to yield P2 that contains membranes and synaptosomes and the cytoplasmic S2. P2 fractions were resuspended in 0.32 m sucrose, which was then layered onto 0.8 m sucrose. After being centrifuged at 9100 × g for 15 min in a swinging bucket rotor, synaptosomes were collected from 0.8 m sucrose layer and concentrated by centrifugation at 20 800 × g for 1 h. To further purify the PSD fractions, synaptosomes in the 0.8 m sucrose solution were mixed with 1/19 volume of buffer D containing 200 mm HEPES (pH 7.0), 20% Triton X-100 and 1.5 m KCl. Samples were centrifuged at 20,800 × g for 45 min using a fixed angle rotor. The resulting pellets were resuspended in buffer E containing 1% Triton X-100 and 75 mm KCl using a Dounce mini-homogenizer and centrifuged again at 20,800 × g for 30 min to yield final pellets (PSD fraction), which were washed with 20 mm HEPES (pH 7.9) and dissolved in 1× SDS-PAGE sample buffer.
Western blotting
Homogenates of ACC tissue were prepared in RIPA buffer containing 50 mm Tris-HCl, pH 7.4, 150 mm NaCl, 2 mm EDTA, 1% sodium deoxycholate, 1% Triton X-100, 1 mm PMSF, 50 mm sodium fluoride, 1 mm sodium vanadate, 1 mm DTT, and protease inhibitors cocktails. PSD fractions were directly dissolved in 1× SDS-PAGE sample buffer. All the protein samples were boiled in 95°C water bath for 10 min before Western blotting. The homogenates were resolved on SDS-PAGE and transferred to nitrocellulose membranes, which were incubated in the TBS buffer containing 0.1% Tween 20 and 5% milk for 1 h at room temperature before incubation with a primary antibody overnight at 4°C. After wash, the membrane was incubated with an HRP-conjugated secondary antibody in the same TBS buffer for 1 h at room temperature. Immunoreactive bands were visualized by ChemiDocTM XRS + Imaging System (Bio-Rad) using enhanced chemiluminescence (Pierce) and analyzed with ImageJ (NIH). The primary antibodies used in the present study were as follows: anti-CaMKIIα (1:2000, Cell Signaling, 11945), anti-p-CaMKIIα Thr286 (1:2000, Sigma, SAB4300228), anti-GluR1 (1:1000, Abcam, ab109450), anti-GluR2 (1:1000, Millipore, MABN71), anti-NR2A (1:1000, Millipore, 04-901), and anti-NR2B (1:1000, Millipore, MAB5778).
Drug administration
For behavioral experiments, D-AP5 (30 mm, 1 µl; Morris et al.,2013) or KN-93 (5 mm, 1 µl; Ebrahimi et al., 2019) was stereotactically injected into the right side of ACC of mice. In order to minimize tissue damage, one stainless guide cannula with 24 gauge were fixed on the right hemisphere of the skull aiming at ACC (coordinates with respect to bregma: AP: +1.0 mm, ML: −0.4 mm, DV: 2.0 mm) using dental cement. The cannula (exposed on the skull surface for 4 mm) was closed with a curved gauge stainless steel injector needle head to prevent blockade. The mice were allowed to recover for at least 3–4 d from the surgery before behavioral tests. CRD was performed before drug microinjection and at 10 min after microinjection as well. The drug concentrations used in the study were based on our pilot study and the references (Morris et al., 2013; L. Jiang et al., 2017). The CNO injection methods include intraperitoneal injection and microinjection in the present study. Intraperitoneal injection of CNO at a dose of 0.33 mg/ml, 0.2 ml was used to investigate whether activation or inhibition of CL neurons affected ACC neural activities and visceral pain responses (Yin et al., 2020). Microinjection of CNO at a dose of 0.33 mg/ml, 1 µl into ACC was used to study whether AP-5 reversed effect of CL terminal activation in ACC on visceral pain responses (Yi-Hua et al., 2021).
Virus injection
Before surgery, the mice were anesthetized using isoflurane. A heated blanket was used to maintain their body temperature at 36°C. A volume of 300-nl virus (depending on the viral titer) was injected using calibrated glass microelectrodes connected to an infusion pump (RWD) at a rate of 30 nl/min. The coordinates were defined as dorsal-ventral (DV) from the brain surface, anterior-posterior (AP) from the bregma, and mediolateral (ML) from the midline. To minimize the brain injury by surgery, one side brain region (i.e., the right brain region) of CL and ACC was selected in the present study.
For retrograde tracing, rAAV-hSyn-EGFP (AAV2/R, 5.73 × 1012 vg/ml, BrainVTA) was injected into the right ACC (AP: +1.0 mm, ML: −0.4 mm, DV: 2.0 mm) of C57BL/6J mice. For CLGlu-ACCGlu tracing, the virus was spread retrogradely to the CL soma to express GFP. After three weeks, the mice were transcardially perfused under isoflurane anesthesia. Brain slices (30 μm) were prepared for tracing GFP or for co-staining with CaMKII antibodies. For anterograde tracing, rAAV-hSyn-EGFP (AAV2/9, 5.12 × 1012 vg/ml, BrainVTA) were injected into the right CL (AP: +1.0 mm, ML: −2.5 mm, DV: 3.75 mm) to allow the virus to spread to the ACC of the C57BL/6J mice.
To label the excitatory neurons of the CL and ACC, the virus under the control of the cell-specific CaMKIIα promoter was used in the present study (Huang et al., 2019; L. Sun et al., 2020). To activate the excitatory neurons in the CON mice, the virus rAAV-CaMKII-hChR2-EYFP (AAV2/9, 1.00 × 1012 vg/ml, GENE) was used. To inhibit the excitatory neurons in the NMD mice, the virus rAAV-CaMKII-eNpHR-EYFP (AAV2/9, 6.27 × 1012 vg/ml, GENE) was used. Three weeks after injection, EYFP was expressed in the CL and ACC, and optogenetic manipulation was performed. In some experiments, the Cre-dependent viruses rAAV-EF1a-Dio-hChR2-EYFP (AAV2/9, 1.58 × 1013 vg/ml, GENE) and rAAV-EF1a-Dio-eNpHR-EYFP (AAV2/9, 1.00 × 1012 vg/ml, GENE) were used for optogenetic manipulation. The rAAV-CaMKII-hM4D(Gi)-mCherry (AAV2/9, 2.44 × 1012 vg/ml, BrainVTA) and rAAV-CaMKII-hM3D(Gq)-mCherry (AAV2/9, 5.40 × 1012 vg/ml, BrainVTA) were used for chemogenetic manipulations three weeks after injection, with intraperitoneal injection of CNO (2.5 mg/kg, BrainVTA) 1 h before the behavior test. The rAAV-CaMKII-EYFP (AAV2/9, 6.76 × 1012 vg/ml, GENE) and rAAV-EF1a-Dio-EYFP (AAV2/9, 2.01 × 1013 vg/ml, GENE) viruses were used as CON. At each location, 300 nl of the virus was injected at a speed of 30 nl/min. Subsequently, the needle was left in place for an additional 10 min and then slowly withdrawn.
Optogenetic manipulations in vivo
An optical fiber (diameter, 200 µm, Newdoon) was implanted into the right CL or ACC area of mice in a stereotaxic apparatus under anesthesia. The coordinates of ACC (AP: +1.0 mm, ML: −0.4 mm, DV: 1.9 mm) and CL (AP: +1.0 mm, ML: −2.5 mm, DV: 3.65 mm). The implant was secured to the animal's skull with dental cement, as described previously (Zhu et al., 2019; Q. Xiao et al., 2021). The implantable fibers were connected to a laser generator using optic fiber sleeves. The delivery of a 20-s pulse of blue (473 nm, 2–5 mW, 20-ms pulses, 10 Hz) or yellow (594 nm, 3–5 mW, constant) light was controlled by STSI-Optogenetics-LED (Alpha Omega; W. Zhou et al., 2019; Zhu et al., 2019). The same stimulus protocol was applied to the mice in the CON group.
Electrophysiology recordings in vivo
For acute extracellular recordings, a single electrode was implanted in the right CL or ACC. The recording electrode was attached to a 16-channel headstage, and the neuronal signals were amplified, filtered at a bandwidth of 300–5000 Hz, and stored using AlphaLab SNR (Alpha Omega). Neuronal firings in the ACC or CL were recorded in the absence and presence of blue or yellow light stimulation (20 s) under CRD stimulation, as described previously (P.A. Zhang et al., 2017, 2018). The firing rates were calculated using NeuroExplorer 4 (Neuralynx).
Brain slice electrophysiology
Mice from the CON and NMD groups (24–28 g) were killed after deep anesthesia with isoflurane. The brain slice electrophysiology was performed as described previously (W. Zhou et al., 2019). The brain was rapidly removed and embedded in 1.6% high-strength agarose (Type I-B, Sigma). The agarose block with brain tissue was placed into 32°C oxygenated artificial CSF (ACSF) with the following composition (in mm): 93 NMDG, 2.5 KCl, 1.2 NaH2PO4, 30 NaHCO3, 20 HEPES, 5 sodium ascorbate, 2 thiourea, 3 sodium pyruvate, 12 NAC, 10 MgSO4, 0.5 CaCl2, and 25 glucose (pH 7.2–7.4, osmolarity: 300–310 mOsm/kg). Coronal slices of right brain (300 µm) were cut using a vibrating microtome (Leica, VT1200S) and placed into solution at 32°C for 30 min to recover. They were then left to adapt to room temperature in oxygenated holding solution with the following composition (in mm): 94 NaCl, 2.5 KCl, 1.2 NaH2PO4, 30 NaHCO3, 20 HEPES, 5 sodium ascorbate, 2 thiourea, 3 sodium pyruvate, 2 MgSO4, 2 CaCl2, 12 NAC, and 25 glucose (pH 7.2–7.4, osmolarity: 300–310 mOsm/kg). After recovery for 30 min, the slices were perfused with ACSF (in mm): 124 NaCl, 2.5 KCl, 1.2 NaH2PO4, 24 NaHCO3, 5 HEPES, 12.5 glucose, 2 MgSO4, and 2 CaCl2 (pH 7.2–7.4, osmolarity: 300–310 mOsm/kg). The perfusion flow rate was ∼2 ml/min. Only two brain slices were used per animal. Only one neuron was recorded in each brain slice.
Neurons were visualized using infrared differential interference contrast video microscopy with a 40× magnification water-immersion objective (BX51WI, Olympus). Whole-cell patch-clamp recordings were obtained from visually identified ACC and CL neurons. Patch electrodes (4–8 MΩ tip resistance) were fabricated from borosilicate glass capillaries using a Flaming/Brown P-97 micropipette puller (Sutter Instruments Co). The internal solution for recording spontaneous EPSCs (sEPSCs) and action potentials contained (in mm): 133 K-gluconate, 8 NaCl, 0.6 EGTA, 10 HEPES, 2 Mg-ATP, and 0.3 Na-GTP (pH 7.2–7.4, osmolarity: 280–290 mOsm/kg). After GΩ seals (usually >4 GΩ) were formed and the whole-cell configuration was obtained, neurons were tested whether the resting membrane potential (RP) was more negative than −50 mV and direct depolarizing current injections evoked action potentials overshooting 0 mV (Yang and Li, 2000). Data were acquired using a Digidata1440A interface, Multiclamp700B amplifier, and pClamp10 software. Data were sampled and filtered at 10 kHz with the Bessel filter of the amplifier. To ensure high-quality recordings, the series resistance (<20 MΩ) was checked using the membrane test function of the pClamp10 software throughout the experiment. The data were stored on a computer and analyzed offline.
Statistical analyses
Animals with incorrect electrode placement, virus microinjections or optic fiber placement were excluded from the analysis. All data are presented as mean values ± SEM, and normality was checked for all data before comparison. Data with a minimal probability that the animal's value deviates from the mean by >2 SDs are defined as abnormal data. Some studies have found that these abnormal data can be eliminated within the scope of experimental error (Asif et al., 2020). The remaining data are analyzed statistically to obtain the results in the graph. Statistical significance was assessed by one-or two-tailed Student's t test, two-sample t test, and one- or two-way ANOVA followed by Sidak's multiple comparisons test using GraphPad Prism software; p < 0.05 was considered statistically significant.
Results
NMD enhanced neuronal activity and excitability of the CL and EMG activity
To investigate whether the CL is involved in NMD-induced chronic visceral hypersensitivity, we compared the expression of c-Fos, an immediate early gene product and marker of neural activity (Greenberg et al., 1986), in the CL of mice. The number of c-Fos expressing cells was dramatically increased in NMD mice after repetitive CRD stimulation (CON, n = 3 slices; NMD, n = 3 slices, p = 0.001, t = 8.601, df = 4, two-sample t test; Fig. 1A,B). Meanwhile, the frequency of sEPSCs for CL neurons in NMD mice was significantly increased when compared with CON mice (CON, n = 6 cells; NMD, n = 7 cells, p = 0.0097, t = 3.12, df = 11, two-sample t test; Fig. 1C), although the amplitude was not significantly altered (CON, n = 6 cells; NMD, n = 7 cells, p = 0.6865, t = 0.4145, df = 11, two-sample t test; Fig. 1C). The firing frequency of neurons in CL slices from the NMD group was also significantly higher than that of the CON group under current-clamp mode in response to intracellular depolarizing current stimulation at 80, 120, and 160 pA (CON, n = 6 cells; NMD, n = 5 cells, 80 pA: p = 0.02;120 pA: p = 0.023;160 pA: p = 0.004, F(1,27) = 27.34, df = 27, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 1D). In addition, the action potential threshold was significantly hyperpolarized (CON, n = 6 cells; NMD, n = 6 cells, p = 0.002, t = 4.174, df = 10, two-sample t test; Fig. 1E), although the RP was not altered after NMD (CON, n = 6 cells; NMD, n = 6 cells, p = 0.2, t = 1.25, df = 10, two-sample t test; Fig. 1F). These results showed that CL neurons were activated and synaptic transmission was promoted in NMD mice with visceral hypersensitivity. To verify the relationship between the CL and visceral pain, we developed an experimental protocol to observe simultaneously the neural firings and EMG activity after CRD stimulation, as shown in Figure 1G. CRD stimulation at 60 mmHg but not 20 mmHg significantly increased both the CL extracellular discharge frequency and the AUC of the EMG activity from NMD mice when compared with CON mice (CON, n = 8 mice; NMD, n = 8 mice, EC recording: p = 0.0098, F(1,14) = 2.592, df = 14; EMG recording: p = 0.0156, F(1,14) = 15.39, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 1H). Since CRD stimulation at 20 mmHg did not cause any significant alteration, this parameter was used as a CON group for the present study thereafter. Further analysis showed that the AUC of the EMG recordings and the discharge frequency of the extracellular recordings were positively correlated (r = 0.5489, p = 0.0341; Fig. 1I). These data strongly indicated that CL is involved in chronic visceral hypersensitivity induced by NMD.
NMD enhances neuronal excitability of the CL and EMG activity. A, The experimental protocol of CRD stimulation. B, c-Fos expression increased after CRD stimulation at the CL region. Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL region (n = 3 slices, **p < 0.01, two-sample t test). Scale bars: 50 µm. C, Representative traces illustrating sEPSCs of a CL neuron from CON and NMD mice. Bar plot showing the significant increase sEPSC frequency (CON, n = 6 cells; NMD, n = 7 cells, **p < 0.01, two-sample t test). D, Representative action potential traces from the CON and NMD groups evoked by 80-, 120-, and 160-pA current stimulation (left). Bar plot illustrating the higher firing frequency of neurons in NMD mice than in CON mice in response to different injected current intensity (right; CON, n = 6 cells; NMD, n = 5 cells, *p < 0.05, **p < 0.01, two-way RM ANOVA). E, Changes in the action potential threshold in CON and NMD mice (CON, n = 6 cells; NMD, n = 6 cells, **p < 0.01, two-sample t test). F, The RP was not altered after NMD (n = 6 cells, p > 0.05, two-sample t test). G, The experimental protocol of extracellular and EMG recording. H, Representative traces of extracellular and EMG recording of the CON and NMD groups. Bar plot showing the action potentials per second of the extracellular recordings and AUC of the EMG recordings at the same distention pressure (CON, n = 8 mice; NMD, n = 8 mice, **p < 0.01, two-way RM ANOVA). I, Correlation analysis showing a positive correlation between EMG and extracellular recordings at 20 and 60 mmHg CRD stimulation (r = 0.5489, p = 0.0341). See also Extended Data Figure 1-1.
To explore whether CL is involved in somatic pain process, we choose a CFA-induced arthritis as a somatic pain model. Consistent with previous reports (Xiang et al., 2019; L. Zhang et al., 2021), CFA injection into the left hind paw significantly increased percentage of paw withdrawal frequency (CON, n = 6 mice; CFA, n = 6 mice, p = 0.0428, t = 2.319, df = 10, two-sample t test; Extended Data Fig. 1-1A). However, CFA did not enhance c-Fos expression in the CL region when compared with CON mice (contralateral, CON, n = 6 slices; CFA, n = 7 slices, p = 0.7656, t = 0.3056, df = 11; Extended Data Fig. 1-1B; ipsilateral, CON = 5 slices, CFA = 5 slices, p = 0.6275, t = 0.5045, df = 8, two-sample t test; Extended Data Fig. 1-1C). In addition, the frequency and amplitude of sEPSCs was not greatly altered (CON, n = 6 cells; CFA, n = 6 cells, Frequency: p = 0.9289, t = 0.09,153, df = 10; Amplitude: p = 0.9268, t = 0.09,419, df = 10, two-sample t test; Extended Data Fig. 1-1D). The firing frequency of neurons in CL slices from the CFA group was not significantly altered when compared with CON group under current-clamp mode in response to intracellular depolarizing current stimulation at 80 and 120 pA (CON, n = 6 cells; CFA, n = 6 cells, 80 pA: p = 0.8152, 120 pA: p = 0.6389, F(1,10) = 11.8, two-way ANOVA RM followed by Sidak's multiple comparisons test; Extended Data Fig. 1-1E). The action potential threshold and RP was not significantly difference (CON, n = 6 cells; CFA, n = 6 cells, AP: p = 0.8240, t = 0.2283, df = 10; RP: p = 0.4070, t = 0.8655, df = 10, two-sample t test; Extended Data Fig. 1-1F). We used AAV-CaMKII-eNpHR-EYFP and AAV-CaMKII-hChR2-EYFP to mark excitatory neurons of CL region in the CFA and NS mice (Extended Data Fig. 1-1G). Optogenetically silencing or activating the CL did not alter the hindpaw pain behaviors of CFA mice (eNpHR, n = 8 mice, p = 0.7356, t = 0.3444, df = 14; hChR2, n = 8 mice, p = 0.6186, t = 0.5092, df = 14; two-sample t test; Extended Data Fig. 1-1H). Similarly, optogenetically silencing or activating the CL did not alter the hindpaw pain sensitivity of NS mice (eNpHR, n = 8 mice, p = 0.3343, t = 1.000, df = 14; hChR2, n = 8 mice, p = 0.6783, t = 0.4237, df = 14; two-sample t test; Extended Data Fig. 1-1I). Furthermore, neither optogenetically silencing nor optogenetically activating the CL terminals in ACC region altered the hindpaw pain sensitivity in CFA and NS mice, respectively (CFA = 8 mice, p = 0.3865, t = 0.8939, df = 14; NS = 8 mice, p = 0.642, t = 0.4752, df = 14, two-sample t test; Extended Data Fig. 1-1J–L). These data suggested that CL is not involved in inflammatory somatic pain induced by CFA, indicating that CL might be a specific brain region for visceral pain.
CL glutamatergic excitatory neurons participated in visceral pain
To explore which type of neurons in the CL were involved in visceral pain, we double-labeled activated c-Fos neurons with the forebrain excitatory neuron marker CaMKII. As shown in Figure 2A, most neurons were co-labeled with CaMKII. Next, we used AAV-CaMKII-eNpHR-EYFP to mark excitatory neurons in NMD mice (Fig. 2B). The results showed that inhibition of CL neurons by yellow-light significantly suppressed AUC induced by 60 mmHg CRD stimulation in NMD mice (eNpHR, n = 8 mice, p = 0.0007, F(1,14) = 6.430, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2C). There was no significant difference before and after yellow-light inhibition in the EYFP group (EYFP, n = 9 mice, p > 0.05, F(1,16) = 34.45, df = 16, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2D). AAV-CaMKII-hChR2-EYFP was injected to mark excitatory neurons in CON mice. The AAV-CaMKII-EYFP virus (EYFP) was used as a negative control, unless otherwise specified. Activation of the CL by blue-light enhanced AUC in CON mice under 60 mmHg CRD stimulation (hChR2, n = 8 mice, p = 0.0023, F(1,14) = 17.51, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2E). There was no significant difference in the EYFP group before and after blue-light activation (EYFP, n = 8 mice, p > 0.05, F(1,14) = 0.9238, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2F).
Excitatory glutamatergic neurons in the CL participate in the visceral pain experience of NMD mice. A, Representative images of c-Fos neurons (green) and CaMKII neurons (red) co-expression in the CL (left) and the percentage co-expressed in the CL (right). Scale bars: 50 µm. White arrows point to representative neurons. B, Schematic of optogenetic experiments in CON and NMD mice (left); timeline of optogenetic experiments (middle); representative images of viral expression in the CL (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. C, Sample traces and summarized data showing the effects of photostimulation (eNpHR, n = 8 mice, ***p < 0.001, two-way RM ANOVA). D, Summarized data showing the effects of photostimulation (EYFP, n = 9 mice, p > 0.05, two-way RM ANOVA). E, Sample traces and summarized data showing the effects of photostimulation (hChR2, n = 8 mice, **p < 0.01, two-way RM ANOVA). F, Summarized data showing the effects of photostimulation (EYFP, n = 8 mice, p > 0.05, two-way RM ANOVA). G, Representative images of c-Fos neurons (red) and glutamatergic neurons (green) co-expression in the CL (left) and the percentage of co-expression in the CL (right). Scale bars: 50 µm. White arrows point to representative neurons. H, Schematic of optogenetic (left) and representative images of viral expression in the CL (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. I, Sample traces and summarized data showing the effects of photostimulation (eNpHR, n = 6 mice, ***p < 0.001, two-way RM ANOVA). J, Summarized data showing the effects of photostimulation (EYFP, n = 5 mice, p > 0.05, two-way RM ANOVA). K, Sample traces and summarized data showing the effects of photostimulation (hChR2, n = 8 mice, ***p < 0.0001, two-way RM ANOVA). L, Summarized data showing the effects of photostimulation (EYFP, n = 6 mice, p > 0.05, two-way RM ANOVA).
To further verify whether the CL glutamatergic neurons regulate visceral pain, glutamatergic transgenic mice were used in the experiments. Immunofluorescence results showed that most activated c-Fos neurons were co-labeled with glutamatergic neurons (Fig. 2G). Next, we used AAV-EF1a-Dio-eNpHR-EYFP to mark glutamate neurons in VGlu-Cre NMD mice (Fig. 2H). The results showed that yellow-light significantly attenuated the AUC evoked by 60 mmHg CRD stimulation in VGlu-Cre NMD mice (eNpHR, n = 6 mice, p = 0.0004, F(1,10) = 14.52, df = 10, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2I). There was no significant difference the EYFP group before and after yellow-light inhibition (EYFP, n = 5 mice, p > 0.05, F(1,8) = 0.3480, df = 8, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2J). AAV-EF1a-Dio-hChR2-EYFP was injected to mark glutamatergic neurons in VGlu-Cre CON mice. Blue-light greatly increased the AUC evoked by 60 mmHg CRD stimulation in VGlu-Cre CON mice (hChR2, n = 8 mice, p = 0.0006, F(1,14) = 29.52, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2K). There was no significant difference in the EYFP group before and after blue-light stimulation (EYFP, n = 6 mice, p > 0.05, F(1,10) = 2.420, df = 10, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 2L). These data indicated that CL glutamatergic neurons are involved in chronic visceral hypersensitivity in NMD mice.
Modulation of the CL neuronal terminals in ACC region altered ACC neural firings and visceral pain responses
To understand the neural circuits of CL glutamatergic neurons in chronic visceral pain, we anatomically mapped the downstream long-range projections of these neurons. AAV9-hSyn-EGFP virus, an anterograde tracer (Guo et al., 2021; Qin et al., 2021), was stereotaxically delivered into the CL of CON mice. Three weeks later, an EGFP+ signal was observed in the ACC region (Fig. 3A). Next, we determined whether the ACC glutamatergic neurons received the direct projection of CL glutamatergic neurons. For monosynaptic anterograde tracing, AAV2/1-CaMKII-Cre was injected into the CL region to label glutamatergic neurons. To recognize and combine with AAV2/1-CaMKII-Cre in ACC, the virus of AAV2/9-EF1a-Dio-EYFP was injected into ACC to label local neurons. Three weeks later, the AAV2/1-CaMKII-Cre transmitted monosynaptically into ACC region, it specifically labeled glutamatergic neurons of ACC. When AAV2/1-CaMKII-Cre combines with AAV2/9-EF1a-Dio-EYFP in ACC, the EYFP expression will start up in glutamatergic neurons of ACC region (Fig. 3B, green). At the same time, immunofluorescence experiments were performed to further verify the type of EYFP label neurons. The results showed that a large number of EYFP were co-located with CaMKII in ACC (Fig. 3B, red). In addition, the AAV2/R-hSyn-EGFP virus, a retrograde tracer, was stereotaxically delivered into the ACC in CON mice. Three weeks later, an EGFP+ signal was observed in the CL region (Fig. 3C). EGFP-labeled neurons were predominantly colocalized with the CaMKII antibody. Statistic results were shown in the Figure 3D, right. The ACC is an important pain integration center in the brain (Zhuo, 2014). In order to further verify the relationship between the CL-ACC neural circuit and chronic visceral pain, in vivo multichannel recording technology, together with optogenetics, was employed. The AAV-CaMKII-eNpHR-EYFP virus was injected into the CL region of NMD mice (Fig. 3E). Three weeks later, the extracellular discharge recording results showed that yellow-light inhibition of the CL neuronal terminals in the ACC regions significantly decreased the frequency of extracellular firing (n = 8 mice, Pre-light vs Light on: p = 0.0037; Light on vs Light off: p = 0.0224, F(2,14) = 8.660, RM one-way ANOVA; Fig. 3F) and attenuated the AUC evoked by 60 mmHg CRD stimulation in NMD mice (eNpHR, n = 11 mice, p = 0.0004, F(1,20) = 18.30, df = 20, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 3G). There was no significant difference in the EYFP group before and after yellow-light inhibition (EYFP, n = 10 mice, p > 0.05, F(1,18) = 0.2387, df = 20, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 3H). The AAV-CaMKII-hChR2-EYFP virus was injected into the CL region of CON mice (Fig. 3I). The extracellular discharge recording experiment was conducted three weeks later. The results showed that blue-light stimulation of the CL neuronal terminals in the ACC regions significantly increased the frequency of extracellular firing (n = 8 mice, Pre-light vs Light on: p = 0.0058; Light on vs Light off: p = 0.0231, F(2,14) = 7.899, RM one-way ANOVA; Fig. 3J) and AUC evoked by the 60 mmHg CRD stimulation in CON mice (hChR2, n = 11 mice, p = 0.0158, F(1,20) = 15.21, df = 20, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 3K). There was no significant difference in the EYFP before and after blue-light stimulation (EYFP, n = 8 mice, p > 0.05, F(1,14) = 10.29, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 3L).
Stimulation of the CL terminal alters ACC activity and the visceral pain experienced by NMD mice. A, Schematic of the anterograde virus tracing strategy. Typical images of AAV9-hSyn-EGFP injection sites within the CL (left) and viral expression in the ACC (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. B, Schematic of the virus tracing strategy and representative images of EYFP neurons (green) and CaMKII neurons (red) co-expression in the ACC. Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. C, Schematic of the retrograde virus tracing strategy. Typical images of AAV2/R-hSyn-EGFP injection sites within the ACC (left) and viral expression in the CL (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. D, Representative images of GFP neurons (green) and CaMKII neurons (red) co-expression in the CL. Scale bars: 50 µm. The percentage of GFP and CaMKII-positive neurons in the CL. White arrows point to representative neurons. E, Schematic of optogenetic and in vivo recording of NMD mice. F, Sample traces (left) and summarized data (right) showed the firing rate of ACC neurons in the NMD mice before, during, and after light photostimulation with multiple channel recordings (n = 8 mice, *p < 0.05, **p < 0.01, one-way ANOVA). G, Summarized data showing the effects of photostimulation (eNpHR, n = 11 mice, ***p < 0.001, two-way RM ANOVA). H, Summarized data showing the effects of photostimulation (EYFP, n = 10 mice, p > 0.05, two-way RM ANOVA). I, Schematic of optogenetic and in vivo recording of CON mice. J, Sample traces (left) and summarized data (right) showing the firing rate of ACC neurons in the CON mice before, during, and after light photostimulation with multiple channel recordings (n = 8 mice, *p < 0.05, **p < 0.01, one-way ANOVA). K, Sample traces and summarized data showing the effects of photostimulation (hChR2, n = 11 mice, *p < 0.05, two-way RM ANOVA). L, Summarized data showing the effects of photostimulation (EYFP, n = 8 mice, p > 0.05, two-way RM ANOVA).
NMD enhanced neuronal excitability of the ACC and EMG activity
Previous results raised the question whether the ACC is involved in NMD-induced chronic visceral hypersensitivity. We compared the expression of c-Fos in the ACC between NMD and CON mice. As shown in Figure 4A, the number of c-Fos expression cells was dramatically increased in NMD mice after repetitive CRD stimulation (CON, n = 4 slices; NMD, n = 4 slices, p = 0.0002, t = 7.842, df = 6, two-sample t test; Fig. 4A). Meanwhile, the patch clamp results showed that the neurons of NMD mice had a significantly higher action potential firing rate when compared with CON mice (CON, n = 6 cells; NMD, n = 8 cells, 100 pA: p = 0.0465, t = 2.406, df = 26, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 4B). The action potential threshold was significantly hyperpolarized in NMD mice when compared with CON mice (CON, n = 8 cells; NMD, n = 8 cells, p = 0.0189, t = 2.654, df = 14, two-sample t test; Fig. 4C). However, the RP was not significantly altered after NMD (CON, n = 8 cells; NMD, n = 8 cells, p = 0.9677, t = 0.04,122, df = 14, two-sample t test; Fig. 4D). Next, the sEPSCs of ACC neurons were compared between the CON and NMD groups. Representative traces from two typical neurons from ACC slices in CON and NMD mice showed an increase in frequency, without obvious changes in the amplitude of sEPSCs. The average values suggested a significant increase in the frequency of sEPSCs in the ACC of the NMD mice (CON, n = 8 cells; NMD, n = 8 cells, Frequency: p = 0.0345, t = 2.342, df = 14; Amplitude: p = 0.0981, t = 1.772, df = 14, two-sample t test; Fig. 4E). These data indicated hyperexcitation of ACC neurons in the NMD group.
NMD enhances neuronal excitability of the ACC and EMG activity. A, The expression of c-Fos at the ACC increased after CRD stimulation (n = 4 slices, ***p < 0.001, two-sample t test). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. B, Representative traces of action potential frequency recorded in the ACC of CON and NMD mice (left). Changes in action potential frequency in CON and NMD mice (right; CON, n = 6 cells, NMD, n = 8 cells, *p < 0.05, two-way RM ANOVA). C, Changes in action potential threshold in CON and NMD mice (n = 8 cells, *p < 0.05, two-sample t test). D, The RP threshold was not significantly altered between CON and NMD mice (n = 8 cells, p > 0.05, two-sample t test). E, Representative traces of sEPSCs recorded in the ACC of CON and NMD mice. Bar plots of the amplitude and frequency of sEPSCs recorded in the ACC of CON and NMD mice (n = 8 cells, *p < 0.05, two-sample t test). F, Representative traces of extracellular recording in the ACC of CON and NMD mice. Bar plots describing the action potentials per second of extracellular recording at the same distention pressure (n = 8 mice, **p < 0.01, two-way RM ANOVA). G, Representative traces of EMG recording of CON and NMD mice. Bar plot describing the AUC of EMG recording at the same distention pressure (n = 8 mice, **p < 0.01, two-way RM ANOVA). H, Correlation analysis showing a positive correlation between the EMG and extracellular recordings in NMD mice at 20 and 60 mmHg CRD stimulation (r = 0.6434, p = 0.0177).
The experimental protocol used to verify the relationship between ACC and visceral pain is the same as shown in Figure 1G. The results showed that the AUC of NMD mice was significantly higher than CON mice under the same CRD stimulation. Meanwhile, the ACC extracellular discharge frequency significantly increased compared with CON mice (CON, n = 8 mice; NMD, n = 8 mice, EC recording: p = 0.0057, F(1,14) = 9.064, df = 14; EMG recording: p = 0.0032, F(1,14) = 10.18, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 4F,G). Further, EMG and extracellular recordings were positively correlated (r = 0.6434, p = 0.0177; Fig. 4H). These data indicated that ACC may be involved in chronic visceral hypersensitivity induced by NMD.
ACC glutamatergic excitatory neurons participated in visceral pain
To explore which type of neurons in the ACC were involved in visceral pain, we double-labeled activated c-Fos neurons with CaMKII. Immunofluorescence results showed that 85% of the activated c-Fos neurons were co-labeled with CaMKII-positive neurons (Fig. 5A). AAV-CaMKII-eNpHR-EYFP was used to mark excitatory neurons in the NMD mice (Fig. 5B). The results showed that yellow-light inhibition of ACC significantly reversed AUC in NMD mice induced by 60 mmHg CRD stimulation (eNpHR, n = 8 mice, p = 0.0486, F(1,14) = 3.753, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5C). As expected, yellow-light stimulation of ACC did not significantly affect AUC in the EYFP group (EYFP, n = 7 mice, p > 0.05, F(1,12) = 2.693, df = 12, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5D). AAV-CaMKII-hChR2-EYFP was injected in CON mice. Conversely, stimulation of the ACC with blue-light significantly increased AUC in the CON mice induced by 60 mmHg CRD (hChR2, n = 7 mice, p = 0.003, F(1,12) = 17.32, df = 12, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5E). There were no significant differences before and after light stimulation of the ACC in the EYFP group (EYFP, n = 10 mice, p > 0.05, F(1,18) = 10.49, df = 18, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5F).
ACC glutamatergic excitatory neurons participate in visceral pain of NMD mice. A, Representative images of c-Fos neurons (green) and CaMKII neurons (red) co-expression in the ACC of NMD mice (left) and the percentage of co-expression in the ACC (right). Scale bars: 50 µm. White arrows point to representative neurons. B, Schematic of optogenetic (left) and representative images of viral expression in the ACC (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the ACC. Scale bars: 50 µm. C, Sample traces and summarized data showing the effects of photostimulation (eNpHR, n = 8 mice, *p < 0.05, two-way RM ANOVA). D, Summarized data showing the effects of photostimulation (EYFP, p > 0.05, n = 7 mice, two-way RM ANOVA). E, Sample traces and summarized data showing the effects of photostimulation (hChR2, n = 7 mice, **p < 0.01, two-way RM ANOVA). F, Summarized data showing the effects of photostimulation (EYFP, n = 10 mice, p > 0.05, two-way RM ANOVA). G, Representative images of c-Fos neurons (red) and glutamatergic neurons (green) co-expression in the ACC of VGlu2-Cre mice (left), and the percentage of co-expression in the ACC (right). Scale bars: 50 µm. White arrows point to representative neurons. H, Schematic of optogenetic (left) and representative images of viral expression in the ACC (right). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. I, Sample traces and summarized data showing the effects of photostimulation (eNpHR, n = 6 mice, *p < 0.05, two-way RM ANOVA). J, Summarized data showing the effects of photostimulation (EYFP, n = 5 mice, p > 0.05, two-way RM ANOVA). K, Sample traces and summarized data showing the effects of photostimulation (hChR2, n = 6 mice, **p < 0.01, two-way RM ANOVA). L, Summarized data showing the effects of photostimulation (EYFP, n = 9 mice, p > 0.05, two-way RM ANOVA).
To further verify whether the ACC glutamatergic neurons regulate visceral pain, glutamatergic transgenic mice were used in the present experiments. Immunofluorescence results showed that 95% of the activated c-Fos neurons were co-labeled with glutamatergic neurons (Fig. 5G). Next, we used AAV-EF1a-Dio-eNpHR-EYFP to mark glutamatergic neurons in the VGlu-Cre NMD mice (Fig. 5H). The results showed that yellow-light significantly reversed AUC induced by 60 mmHg CRD stimulation in VGlu-Cre NMD mice (eNpHR, n = 6 mice, p = 0.0128, F(1,10) = 17.63, df = 10, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5I). There were no significant differences before and after yellow-light in the EYFP group (EYFP, n = 5 mice, p > 0.05, F(1,8) = 33.57, df = 8, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5J). AAV-EF1a-Dio-hChR2-EYFP was injected to mark glutamatergic neurons in VGlu-Cre CON mice. Stimulating the CL with blue-light significantly increased AUC induced by 60 mmHg CRD stimulation in VGlu-Cre CON mice (hChR2, n = 6 mice, p = 0.003, F(1,10) = 11.67, df = 10, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5K). There were no significant differences before and after blue-light in the EYFP group (EYFP, n = 9 mice, p > 0.05, F(1,16) = 9.045, df = 16, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 5L). These data indicated that ACC glutamatergic neurons are indeed involved in mediating chronic visceral hypersensitivity in NMD mice.
Modulation of ACC activity altered CL-mediated visceral pain
Next, we determined whether the CLGlu-ACCGlu pathway regulates chronic visceral pain. Three weeks after the injection of AAV-CaMKII-hChR2-EYFP into the CL and AAV-CaMKII-hM4D(Gi)-mcherry into the ACC in CON mice, EMG was performed after CRD stimulation (Fig. 6A). The virus expression was shown in the Figure 6B. Under 60 mmHg CRD stimulation, blue-light activation of the CL terminal in the ACC region significantly increased the AUC of CON mice, which was reversed by chemogenetic inhibition of ACC glutamatergic neurons (CON, n = 8 mice, Light off vs Light on: p = 0.0261; Light on vs Light on+CNO-hM4D: p = 0.0483, F(1,14) = 4.029, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 6C,D). Similarly, AAV-CaMKII-eNpHR-EYFP and AAV-CaMKII-hM3D(Gq)-mcherry were injected into the CL and ACC, respectively, in NMD mice (Fig. 6E). The virus expression showed as the figure (Fig. 6F). Under 60 mmHg CRD stimulation, yellow-light inhibition of the CL terminal in the ACC region significantly reduced the AUC of NMD mice, which was reversed by chemogenetic activation of ACC glutamatergic neurons (NMD, n = 6 mice, Light on vs Light off: p = 0.0081; Light on vs Light on +CNO-hM3D: p = 0.0422, F(1,10) = 13.61, df = 10, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 6G,H). These data suggest that modulation of ACC glutamatergic neuron activity affects visceral pain formation induced by CL glutamatergic neuron activity.
ACC regulation affects CL-mediated chronic visceral pain in mice. A, Schematic of optogenetic and chemogenetic protocol. B, Representative images of viral expression in the CL (top) and ACC (bottom). Scale bars: 20 µm. The white boxes depict the area shown in the box of the ACC. Scale bars: 50 µm. C, The sample traces of EMG. D, Summarized data showing the effects of photostimulation. Blue-light stimulation of the ACC can increase chronic visceral pain, and chemogenetic stimulation can reverse chronic visceral pain, in CON mice by inhibition the excitability of neurons in the ACC (n = 8 mice, *p < 0.05, two-way RM ANOVA). E, Schematic of optogenetic and chemogenetic protocol. F, Representative images of viral expression in the CL (top) and ACC (bottom). Scale bars: 20 µm. The white boxes depict the area shown in the box of the ACC. Scale bars: 50 µm. G, The sample traces of EMG. H, Summarized data showing the effects of photostimulation. Yellow-light stimulation of the ACC can significantly decrease chronic visceral pain, and chemogenetic stimulation can reverse chronic visceral pain, in NMD mice by activation the excitability of neurons in the ACC (n = 6 mice, *p < 0.05, **p < 0.01, two-way RM ANOVA).
Alteration in NMDAR subunit expression and function in ACC of NMD mice
Finally, we investigated the molecular mechanisms underlying chronic visceral hypersensitivity in the NMD mice. The NMDAR and AMPA receptor (AMPAR) trafficking to PSD are considered the important mechanism for central sensitization of chronic pain (X.H. Xu et al., 2020; H. Jiang et al., 2021). For this reason, we investigated whether the synaptic distribution of NMDAR or AMPAR is altered in NMD mice. We purified the PSD fraction from ACC of CON and NMD mice and performed Western blottings on their lysates. As shown in Figure 7A, the NR2A and NR2B subunits of the NMDAR in the PSD fraction were significantly higher in the NMD mice than CON (CON, n = 3 mice, NMD, n = 3, p = 0.001, t = 8.586, df = 4; p = 0.0024, t = 6.805, df = 4, two-sample t test; Fig. 7B). We did not find changes in the PSD distribution of the AMPAR subunits GluR1 and GluR2 between the CON and NMD mice ACC (CON, n = 3 mice, NMD, n = 3 mice, p > 0.05, two-sample t test; Fig. 7B). Our results suggest that the alteration of NMDAR subunits in the PSD fraction are responsible for the central sensitization of chronic visceral hypersensitivity in the NMD mice. In support of this hypothesis was the finding that activation of CaMKIIα, a protein kinase critical for neuronal plasticity and downstream of NMDAR (G.Y. Xu and Huang, 2004; Lisman et al., 2012), was hyperactive in the NMD mice ACC compared with CON (CON, n = 3 mice, NMD, n = 3 mice, p = 0.0003, t = 12.31, df = 4; p > 0.05, two-sample t test; Fig. 7B).
Alteration of NMDAR subunit expression and function in the PSD of ACC of NMD mice. A, Western blottings showing increased NR2A and NR2B subunits of NMDAR and enhanced CaMKIIα activation in the PSD of the NMD mice ACC, compared with CON mice ACC. B, Quantification of NR2A/PSD95 as in panel A (n = 3 mice, **p < 0.01, two-sample t test). Quantification of NR2B/PSD95 as in panel A (n = 3 mice, **p < 0.01, two-sample t test). Quantification of GluR1/PSD95 as in panel A (n = 3 mice, p > 0.05, two-sample t test). Quantification of GluR2/PSD95 as in panel A (n = 3 mice, p > 0.05, two-sample t test). Quantification of p-CaMKIIα/CaMKIIα normalized to CON mice ACC as in panel A (n = 3 mice, ***p < 0.001, two-sample t test). Quantification of CaMKIIα/PSD95 as in panel A (n = 3 mice, p > 0.05, two-sample t test). C, D-AP5 or KN-93 increased visceral pain threshold in NMD mice (n = 8 mice, **p < 0.01, ***p < 0.001, two-sample t test). D, Schematic of chemogenetic and D-AP5 injected protocol in CON mice. Sample traces and summarized data showing the effects of chemogenetic activation the CL neuron terminal in the ACC and D-AP5 inhibition the NMDAR in the ACC (n = 8 mice, **p < 0.01, two-way RM ANOVA).
To determine the functional roles of NMDAR and CaMKII, NMDAR antagonist (D-AP5) or CaMKII inhibitor (KN-93) was injected into the ACC region through the cannula. D-AP5 or KN-93 significantly enhanced distention threshold of NMD mice (n = 8 mice, NMD vs NMD+D-AP5: p = 0.008, t = 3.092, df = 14; NMD vs NMD+KN-93: p = 0.0006, t = 4.444, df = 14, two-sample t test; Fig. 7C). Next, we explored whether NMDARs were involved in CL-to ACC circuit. Three weeks after the AAV-CaMKII-hM3D(Gq)-mcherry was injected into the CL region of CON mice, the EMG recordings were performed after CRD stimulation. The results showed that CNO activation of the CL glutamatergic neuron terminals in the ACC region significantly enhanced the AUC under 60 mmHg CRD stimulation, which was reversed by D-AP5 inhibition of the NMDAR in the ACC (n = 8 mice, Pre vs CNO-hM3D: p = 0.0012; CNO-hM3D vs CNO-hM3D+D-AP5: p = 0.0096, F(1,14) = 34.26, df = 14, two-way RM ANOVA followed by Sidak's multiple comparisons test; Fig. 7D). Together, these data suggested that excessive NMDAR subunits and overactive CaMKIIα in the ACC PSD region likely accounts for the central sensitization of chronic visceral hypersensitivity in the NMD mice.
Discussion
Chronic visceral pain is an uncomfortable symptom in patients with IBS. Effective treatment options for this disease are limited. The lack of basic research on the mechanisms underlying visceral pain lags behind the development of effective drugs or strategies. Recently, many efforts have focused on the investigation of central nervous mechanisms. Identification of the central neural circuitry underlying chronic visceral pain has been a major challenge and key step in its treatment. A novel finding of the present study is the identification of the CL-ACC pathway as an important neural circuit involved in the development of chronic visceral pain in an adult mouse model of IBS induced by NMD. In particular, enhanced CL glutamatergic projection activities facilitate glutamatergic neuron activation in the ACC region, eventually contributing to chronic visceral pain.
The CL is a distinct center for visceral pain but not for CFA-evoked inflammatory pain process
While the CL has been reported to participate in consciousness (Chau et al., 2015; Torgerson et al., 2015; Smith et al., 2019) and the spatiotemporal coordination of slow-wave activity (Narikiyo et al., 2020), it remains unclear whether it regulates visceral pain. Since NMD is a well-established chronic visceral hypersensitivity model in adult mice (Z.Y. Chen et al., 2021), we therefore used this model to explore the central mechanisms. The present study claims that CL participates in the regulation of chronic visceral pain. CRD stimulation induced a dramatic increase in c-Fos expression in the CL of NMD mice when compared with that in the CON group (Fig. 1B), indicating that CL could be activated by visceral pain stimulation. This conclusion was further confirmed by in vivo and in vitro electrophysiological recordings, which indicated that CRD stimulation significantly increased CL neural activation and synaptic transmission compared with CON mice. Notably, we developed a novel technology to simultaneously record the EMG activities of abdominal oblique muscle and the extracellular discharges of CL neurons after CRD stimulation (Fig. 1G). This approaches greatly help us to study the correlation between neural firings of brain region and the muscle contraction, which is a well-developed objective index of visceral pain (Y.Y. Zhou et al., 2012; H.H. Zhang et al., 2013). Indeed, the EMG data were positively correlated with neural firing frequency after painful CRD stimulation (i.e., 60 mmHg). Non-noxious CRD stimulation (20 mmHg) did not produce any effect and was used as a control (Fig. 1H). More importantly, optogenetic enhancement of CL activity produced visceral pain sensitization, while optogenetic suppression of CL activity attenuated visceral pain sensitization (Fig. 2C,E). These data suggest that the CL is sufficient and necessary for the experience of visceral pain. A very surprising finding is that von Frey fiber stimulation failed to induce a significant increase in c-Fos expression in the CL region in a mouse model of somatic pain model induced by CFA injection (Extended Data Fig. 1-1B,C). Moreover, CFA injection did not alter the excitability and synaptic transmission of CL neurons in CFA-induced somatic pain mice (Extended Data Fig. 1-1D–F). Together, the present results indicated that the CL might be involved in visceral pain rather than somatic pain. But this conclusion needs a further verification.
Glutamatergic neurons in the CL region mediate visceral pain
One important finding of the present study was that glutamatergic neurons in the CL mediate visceral pain. As reported previously, the CL region contains many types of neurons, including glutamatergic, GABAergic, and parvalbumin neurons (Rahman and Baizer, 2007; Baizer et al., 2020; Pirone et al., 2020). To determine the type of CL neurons that participate in the visceral pain process, multiple techniques including optogenetics and viral tracing were employed. We showed for the first time that the majority of c-Fos-positive neurons activated by CRD stimulation were CaMKIIα-positive neurons (Fig. 2A). This is consistent with previous reports that the majority of c-Fos positive cells were glutamatergic neurons in the CL region (Borroto-Escuela and Fuxe, 2020). Importantly, the present study provided sufficient evidence indicating that CL glutamatergic neurons were specifically activated by CRD stimulation. Although it is unclear whether CL GABAergic neurons are activated by CRD stimulation, CL glutamatergic neurons serve as one of the key elements in regulating visceral pain. This suggests the possibility of a neuron-type-specific drug delivery strategy for the treatment of visceral pain in the brain.
CL Glutamatergic projection to ACC glutamatergic neurons regulates visceral pain
Another important finding was the identification of the direct CLGlu-ACCGlu neural circuit, which uniquely mediates visceral pain. Using a dual-virus strategy, we confirmed a direct neural connection between the CL and ACC regions (Fig. 3A,C). This is in agreement with a previous report indicating that glutamatergic neurons in the CL region directly project to the ACC region (Narikiyo et al., 2020). However, the previous report did not identify the neural types involved in this circuit. We provide novel evidence of a direct glutamatergic-glutamatergic connection between these two brain regions by epigenetic modulation of CL glutamatergic terminals in ACC region (Fig. 3B). In addition, we showed that the CLGlu-ACCGlu neural circuit is exclusively involved in visceral pain processing. Moreover, we demonstrated that this pathway was necessary and sufficient for visceral pain. This conclusion is based on the following observations. First, glutamatergic neurons in both the CL and ACC regions were activated by CRD stimulation and were shown to play a role in regulating visceral pain. Second, optogenetic activation of the CL-ACC glutamatergic pathway produced visceral pain, whereas optogenetic inhibition attenuated visceral pain (Fig. 3E–L). Third, the regulatory effect of CL glutamatergic neuron activity on visceral pain was almost completely blocked by the activity of ACC glutamatergic neurons (Fig. 6A–H). ACC has been reported to modulate both sensory and negative affective components of chronic pain (Y.G. Sun et al., 2008; X. Xiao and Zhang, 2018). Immunofluorescence and in vitro electrophysiological results indicated that ACC glutamatergic neurons might be involved in visceral pain. The ACC receives inputs from both CFA-evoked inflammatory and visceral information. One question that remains to be answered is how the ACC distinguishes between CFA-evoked inflammatory and visceral pain inputs. Although it is very difficult to answer this question, our results provide a novel idea that the CL might be a crucial brain region that regulates visceral pain. Of note is that the unilateral brain region was chosen for virus injection and optogenetic modulation in the present study. It is reasonable to hypothesize that bilateral brain region modulation might produce better effects. The reasons why we chose unilateral brain modulation protocol are mainly based on the following consideration: to minimize the surgery injury and our preliminary experiment results that showed an obvious effect of manipulating right brain region on visceral pain, and the previous studies that regulating unilateral brain regions can change pain responses (Tao et al., 2019; Zhao et al., 2021), Therefore, we decided to do unilateral brain craniotomy to regulate the right CL and ACC function. To avoid effects of the female hormones on pain response, only male mice were used in the present study. This also raises a question on whether the same results are observed in female animals. Therefore, it is worthy of further study in the future.
Alteration of NMDAR subunits in NMD mice ACC
Although the mechanism by which ACC neurons are activated by stimulation of CL glutamatergic presynaptic neurons has not yet to be investigated, our results suggest that excessive NMDAR subunits and overactive CaMKIIα in the ACC PSD region accounts for the central sensitization of CLGlu-ACCGlu neural circuit of NMD mice with chronic visceral hypersensitivity. In support of this hypothesis are previous studies showing that chronic postischemic pain mediates the expression of NMDARs in rat brain (X.H. Xu et al., 2020). We studied the molecular mechanism of the ACC but not the CL brain area because the ACC is the target brain area of CL projections, and to study the distribution of receptors in the PSD rather than the total cell membrane, because the PSD is the terminal location of the neural projections. We detected the distribution of glutamate receptors is because CLGlu-ACCGlu regulates visceral pain, and glutamate receptors are important terminals that control neuronal excitability, but we do not exclude other receptors in the regulation of visceral pain in the CLGlu-ACCGlu, such as GABAAR, TRPV channels and P2X/Y receptors (G.Y. Xu et al., 2008; Lapointe et al., 2015; Huang et al., 2021). To support our hypothesis of NMDAR augmentation, we further demonstrated the enhanced activity of CaMKIIα kinase downstream of NMDAR. However, further studies are needed to determine whether the increased NMDAR in the PSD region is because of increased NMDAR trafficking or increased total NMDRA protein synthesis. Although the potential mechanisms underlying the observed changes in intrinsic membrane excitability in the CL and ACC neurons after NMD have not yet been investigated in details, it is likely that NMD might alter the intrinsic expression of voltage-gated or voltage-independent ion channels within CL or ACC neurons in addition to the observed changes in synaptic transmission. Excitingly, our results suggest that NDMAR inhibitor targeting ACCGlu could treat severe chronic visceral pain. Although optogenetic and chemogenetic studies have shown that glutamatergic neurons play an important role in the circuit, it is difficult to exclude the roles of other neuron types. The detailed neural circuitry among these regions requires further investigation.
In summary, the present study confirmed the presence of a direct glutamatergic pathway between the CL-ACC brain regions, which might serve as a central neural circuit underlying the chronic visceral pain. The detailed working model is shown in Figure 8. Although the ACC receives both somatic and visceral sensory information, the CL maybe receives visceral pain inputs, which helps the ACC to perceive different information. In addition, excessive NMDAR subunits and overactive CaMKIIα in the ACC PSD region might account for the central sensitization of CLGlu-ACCGlu neural circuit of NMD mice with chronic visceral hypersensitivity. These findings reveal the neural circuity of central sensitization underlying chronic visceral pain and might provide a new treatment direction for chronic visceral pain in patients with IBS.
A working model showing that a direct CLGlu-ACCGlu neural circuit mediates chronic visceral pain. A, CRD induces firings of both CL and ACC neurons and visceral pain sensation of CON mice. Blue-light activation of CL glutamatergic neuron terminals in ACC region evoked a lot of more firings of ACC neurons accompanied by an enhanced visceral pain response, which was reversed by chemogenetic inhibition of ACC glutamatergic neurons of CON mice. B, CRD induces a lot of more firings of both CL and ACC neurons and an enhanced visceral pain response of NMD mice when compared with CON mice. Yellow-light inhibition of CL glutamatergic neuron terminals in ACC region suppressed both the ACC neural activity and the visceral hypersensitivity of NMD mice, which was reversed by chemogenetic activation of ACC glutamatergic neurons. 
: indicating a neuron under physiologic condition; 
: indicating a neuron under hyperactive condition.
Extended Data Figure 1-1
CFA does not increase c-Fos expression, neural excitability, and synaptic transmission of the CL, and the effects of optogenetically modulating of CL on visceral pain responses. Connect to Figure 1. A, The left hindpaw of a mouse at 3 d after CFA injection and NS. CFA induced changes in the paw withdrawal frequency (n = 6 mice, *p < 0.05, two-sample t test). B, The expression of c-Fos in the contralateral CL did not increase after CFA injection into the left paw (NS, n = 6 slices; CFA, n = 7 slices, p > 0.05, two-sample t test). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. C, The expression of c-Fos in the ipsilateral CL did not increase after CFA injection into the left paw (NS, n = 5 slices; CFA, n = 5 slices, p > 0.05, two-sample t test). Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. D, Representative traces of sEPSCs recorded in the CL of NS and CFA mice. Bar plots of the frequency and amplitude of sEPSCs recorded in the CL of NS and CFA mice (n = 6 cells, p > 0.05, two-sample t test). E, Representative traces of action potential frequency recorded in the ACC of NS and CFA mice. Bar plots of the action potential frequency in NS and CFA mice (n = 6 cells, p > 0.05, two-way RM ANOVA). F, Changes in action potential threshold and RP threshold were not significantly difference in NS and CFA mice (n = 6 cells, p > 0.05, two-sample t test). G, Schematic of optogenetic experiments in CFA and NS mice and timeline of optogenetic experiments. Representative images of viral expression in the CL. Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. H, Summarized data showing the effects of photostimulation in CFA mice (eNpHR, n = 8 mice, p > 0.05; hChR2, n = 8 mice, p > 0.05, two-sample t test). I, Summarized data showing the effects of photostimulation in NS mice (eNpHR, n = 8 mice, p > 0.05; hChR2, n = 8 mice, p > 0.05, two-sample t test). J, Schematic of optogenetic experiments in CFA and NS mice and timeline of optogenetic experiments. Representative images of viral expression in the CL. Scale bars: 20 µm. The white boxes depict the area shown in the box of the CL. Scale bars: 50 µm. K, Summarized data showing the effects of photostimulation in CFA mice (eNpHR, n = 8 mice, p > 0.05, two-sample t test). L, Summarized data showing the effects of photostimulation in NS mice (hChR2, n = 8 mice, p > 0.05, two-sample t test). Download Figure 1-1, TIF file.
Footnotes
This work was supported by National Natural Science Foundation of China Grants 31730040 and 81920108016 and the Priority Academic Program Development of Jiangsu Higher Education Institutions of China.
The authors declare no competing financial interests.
- Correspondence should be addressed to Guang-Yin Xu at guangyinxu{at}suda.edu.cn or Rui Li at 13771725877{at}163.com
