A total of 95 healthy male C57BL/6J mice (age 6–8 weeks) without physical abnormalities and weighing 20–22 g [Beijing Vital River Laboratory Animal Technology Co., Beijing, China. license number: SCXK (Jing) 2021-006] were used in this study. The mice were allocated as follows: 49 for pain behavioral assays, 15 for immunofluorescence, and 31 for electrophysiological recordings. Cohorts did not overlap across assays. The animals were maintained under controlled environmental conditions with a 12-h light/dark cycle and a temperature of 24 $\pm$ 2 °C. Standard laboratory chow and water were provided ad libitum. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Institute of Acupuncture and Moxibustion, China Academy of Chinese Medical Sciences (approval no. D2024-01-26-09) and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. After data collection, animals were euthanized with an overdose of inhaled isoflurane (R510-22-10, RWD Life Science Co., Shenzhen, Guangdong, China).

For the CFA-treated group, mice received an intraplantar injection of 20 µL of CFA (F5881, Sigma-Aldrich, St. Louis, MO, USA) into the left hindpaw during isoflurane (R510-22-10, RWD Life Science Co.) inhalation anesthesia (1%–4%). The CFA solution was administered slowly and allowed to remain in situ for 2 minutes before withdrawal to ensure complete absorption, with no observable leakage detected. Mice that failed to develop CFA-induced paw inflammation (swelling/redness) were excluded (Supplementary Fig. 1). For the control group, naïve mice were used without sham handling or injection. For all experiments, mice were allocated to experimental groups using a computer-generated random number sequence upon completion of acclimatization. For behavioral cohorts, this randomization was performed after baseline behavior testing to ensure equivalent starting scores across groups.

Gait analysis was performed using the CatWalk XT system (version 10.6, Noldus Information Technology, Wageningen, The Netherlands) to observe spontaneous pain behaviors as previously described [13]. Briefly, the system consists of an enclosed walkway with a glass plate, illuminated by a fluorescent lamp that emits light inside the glass plate. Each mouse was placed individually in the CatWalk system and allowed to walk freely and traverse from one side of the walkway glass plate to the other. Where the mouse paws made contact with the glass plate, light was reflected down and the illuminated contact areas were recorded with a high-speed color video camera positioned underneath the glass plate. The recordings were carried out when the room was completely dark, except for light from the computer screen. The software automatically labeled all areas containing pixels above the set threshold and assigned them to the respective paws. Mice were trained to traverse the walkway for 2 days before baseline measurements. Animals were excluded if they failed to run voluntarily. Each mouse was tested at least three times. Five parameters were analyzed: (1) stand phase: duration in seconds of paw contact with the glass plate in a step cycle; (2) swing phase: duration in seconds of no paw contact with the glass plate in a step cycle; (3) duty cycle: stand duration as a percentage of the step cycle duration [stand / (stand + swing) $\times$ 100]; (4) single stance: duration in seconds of ground contact for a single paw; and (5) max contact max intensity: the maximum intensity of a paw at maximum paw-floor contact. These values were normalized by calculating the ratio between the ipsilateral (affected left hindpaw) and contralateral (unaffected right hindpaw) measurements to quantify the inflammatory response-induced alterations. Animals showing abnormal gait parameters before CFA injection were excluded (Supplementary Fig. 1).

To distinguish between inflammatory pain-induced gait alterations and potential motor deficits, motor coordination and balance were evaluated using the accelerating Rotarod test. Prior to testing, the mice underwent a 2-day habituation protocol on the Rotarod apparatus (Mouse-47650, Ugo Basile, Gemonio, Italy). The assessment was performed 24 h post CFA injection, with the Rotarod accelerating linearly from 0 to 40 rpm over a 300-s period. Three trials were conducted for each mouse, and both the latency to fall and terminal velocity were recorded.

In addition to gait analysis, spontaneous nociceptive behaviors were evaluated using a weight-bearing test system (model BIO-SWB-TOUCH-M, Bioseb, Vitrolles, France). Each mouse was placed in an acrylic glass chamber where independent force plates measured the weight distribution between hindpaws. After a 2-day habituation period, weight-bearing measurements were obtained by averaging the applied force (g) over a 10-s sampling period. The weight-bearing differential was calculated by subtracting the contralateral (right hindpaw) from the ipsilateral (CFA-injected left hindpaw) weight-bearing capacity. The final values represent the mean of six consecutive measurements. Animals showing imbalanced bilateral weight-bearing capacity before CFA injection were excluded (Supplementary Fig. 1).

Mechanical hypersensitivity was assessed using calibrated von Frey filaments (37450-275, Ugo Basile). Mice were individually housed in transparent glass chambers positioned on an elevated mesh platform. Prior to testing, animals underwent a habituation protocol consisting of 1-h sessions for 2 consecutive days, followed by a 30-min acclimation period before each test session. The mechanical withdrawal threshold was determined by applying von Frey filaments perpendicularly to the midplantar surface of the ipsilateral hindpaw with sufficient force to cause filament bending for 3–5 s. A response was considered positive if the animal exhibited any nocifensive behaviors, including brisk paw withdrawal, licking, or shaking of the paw, either during application of the stimulus or immediately after the filament was removed. Paw withdrawal thresholds were quantified using the modified Dixon up-down method [14, 15]. The 50% threshold was calculated according to the formula: 50% threshold (g) = 10 ^{(X + kd)}/10⁴, where X represents the log unit value of the final filament, k is the tabular value for the pattern of positive/negative responses, and d denotes the mean difference in log units between filaments. Animals exhibiting abnormal mechanical sensitivity (withdrawal threshold 0.5 g or $>$1.5 g) under normal conditions were excluded (Supplementary Fig. 1).

Baseline measurements for gait parameters, weight-bearing distribution, and mechanical sensitivity were obtained 24 h prior to CFA (F5881, Sigma-Aldrich) injection, with subsequent assessments conducted on days 1, 3, 5 and 7 post injection. Motor coordination was evaluated using the Rotarod test. This was conducted only at the 24 h post-injection timepoint. All behavioral tests were performed at approximately 09:00 AM.

Mice were deeply anesthetized with urethane (1.5 g/kg, U2500, Sigma-Aldrich) followed by transcardial perfusion with saline. It should be noted that no additional acute stimulation was conducted before perfusion. The lumbar spinal cord enlargements were harvested and fixed in 4% paraformaldehyde (PFA, P1110, Solarbio, Beijing, China), then dehydrated in 30% sucrose until the tissue sank to the bottom. After 24 h, the tissue was embedded in an optimal cutting temperature compound (4583, Sakura Finetek, Torrance, CA, USA), frozen, and sectioned at 30-µm thickness using a cryostat (CM1860, Leica, Wetzlar, Germany). The sections were collected in PBS-filled wells. Following PBS washing, the sections were blocked with 3% donkey serum (BMS0140, Abbkine, Wuhan, Hubei, China) and 0.5% Triton X-100 (T9284, Sigma-Aldrich) in PBS for 45 min at room temperature (RT). The sections were then incubated overnight at 4 °C in PBS containing 1% donkey serum and 0.5% Triton X-100, and the following primary antibodies: anti- isolectin B4 (IB4)-488 (1:500, I21411, Invitrogen, Carlsbad, CA, USA), mouse anti-protein kinase C-gamma (PKC$\gamma$) (1:300, 13-3800, Invitrogen), and rabbit anti-c-Fos (1:1000, ABE457, Millipore, Burlington, MA, USA). After three PBS washes, sections were incubated for 1.5 h at RT with secondary antibodies (donkey anti-mouse Alexa Fluor 647, 1:1000, ab150111, Abcam, Cambridge, UK; donkey anti-rabbit Alexa Fluor 555, 1:1000, ab150062, Abcam) and DAPI (1:50,000, D3571, Invitrogen) in PBS containing 1% donkey serum and 0.5% Triton X-100. Following the PBS washes, the sections were mounted on silane-coated slides and visualized using a confocal microscopy system (Nikon AX, Nikon Instruments Inc., Tokyo, Japan). An experienced researcher performed the quantification of c-Fos-positive cells, with the placement of the region of interest (ROI) borders illustrated in Supplementary Fig. 2.

In vivo extracellular recordings were performed from SDH neurons in L4–L6 segments. Mice were anesthetized with intraperitoneal urethane (1.5 g/kg, U2500, Sigma-Aldrich). A laminectomy was performed at T13 and L1 vertebrae to expose the lumbar enlargement (L4–L6). Following careful removal of the dura mater, the spine was immobilized using a stereotaxic frame (SR-6R-HT, Narishige, Tokyo, Japan) to minimize movement artifacts. The exposed spinal cord was continuously perfused with warm physiological saline. Core body temperature was maintained using a heating pad. Neural activity was recorded using a 32-channel silicon probe (A1$\times$32-Poly2-5mm-50s-177, Neuronexus, Ann Arbor, MI, USA). This was inserted perpendicularly into the lateral hindpaw region of the dorsal horn at a depth of approximately 850 µm from the dorsal surface, and 300 µm lateral to the central vessel. The insertion procedure was carried out under a stereoscope (SZ61, Olympus, Tokyo, Japan) to confirm the spinal surface (defined as zero). A micromanipulator (DMA-1510, Narishige) was employed to guide insertion depth and lateral distance. The array was attached to the headstage. Signals were processed through a preamplifier (LB-0164-1, Blackrock Microsystems, Salt Lake City, UT, USA) with a bandpass filter of 250 Hz–5 kHz, digitized, and acquired using a neuronal signal acquisition system (Cerebus-128, Cyberkinetics, Foxborough, MA, USA). Offline spike sorting and analysis were performed using Spike2 software (version 10.01a, Cambridge Electronic Design, Cambridge, UK). The quality control criteria for sorted units were as follows: (i) units were required to be well-isolated from noise and other units, quantified by an Isolation Distance $>$15 and an L-ratio 0.1; (ii) units had a refractory period violation rate of $\le$2%; (iii) the signal-to-noise ratio (SNR) was required to be $\ge$3; (iv) units exhibited stable waveforms throughout the recording session, with no evidence of continuous monotonic drift; and (v) waveform shape was consistent across different stimulus conditions. Representative waveforms and autocorrelogram of a sorted single unit were shown in Supplementary Fig. 3.

The receptive fields (RFs) of recorded neurons were identified through mechanical probing of the ipsilateral hindpaw. Mechanical stimulation of the RFs consisted of brush stimulation using a camel-hair brush. Calibrated force was applied using an electronic rodent pincher (BIO-RPM, Bioseb) consisting of a pair of blunt forceps. One arm of the forceps was flat, while the opposing arm was tipped with a spherical cap 3.5 mm in diameter and 1 mm in height with smooth edges. For brush stimuli, the center of the hindpaw RF was stroked in the rostral to caudal direction about once per second for 15 s. For touch and pinch stimuli, the spherical cap was placed at the center of the hindpaw RFs. The applied force was then quickly increased by hand until it reached 40 g for touch stimulus and 150 g for pinch stimulus. This was held for 15 s using a timer. Stimuli were given in a predefined order (brush, 40 g touch, followed by 150 g pinch) in order to avoid potential diffuse noxious inhibitory controls (DNIC) elicited by the 150 g pinch. Each type of stimulus was given 2–3 times to improve reliability. The inter-trial interval was approximately 1–2 min to avoid neuronal fatigue in response to repeated stimuli. A 10-s stable response period (Supplementary Fig. 4) from each 15-s mechanical stimulus was analyzed to minimize artifacts from potential hand movements. The spontaneous firing rates were defined as the mean frequency of a 30-s baseline period before any stimulation. The stimulus-evoked firing rates were defined as the mean frequency of the 10-s stable response window minus the mean frequency of the 10-s pre-stimulus baseline. Recordings containing fewer than three neurons per animal were excluded to ensure an adequate sample size for population-level analysis (Supplementary Fig. 1).

Neurons were classified into three distinct categories based on their response properties to mechanical stimulation as described previously [16]: (1) LTM neurons that responded exclusively to innocuous stimuli (dynamic brush and/or static touch); (2) WDR neurons that exhibited graded responses to both innocuous and noxious stimuli (pinch), with greater responses to the latter; and (3) NS neurons that responded selectively to noxious pinch.

The depth of each probe was calculated based on the distance from the electrode tip to the spinal cord surface and the spacing between probes. The depth bin for analysis was set as 25 µm. It should be noted that tissue edema caused by surgery and exposure of the spinal cord beyond control may alter the estimated depths. Other potential limitations such as the tilt of the electrode tip and zeroing deviation were controlled as described in the insertion procedure above.

All data are presented as the mean $\pm$ SEM. Statistical analyses were performed using GraphPad Prism (version 9.0, GraphPad Software Inc., San Diego, CA, USA) and SPSS (version 26.0, IBM Corp., Armonk, NY, USA). The Shapiro-Wilk test was used to assess the normality of data distribution (SPSS 26.0). For normally distributed data, comparisons between two groups were made using the unpaired Student’s t-test (SPSS 26.0). Time course comparisons between groups were analyzed using two-way ANOVA followed by Šídák’s multiple comparisons test (GraphPad Prism 9.0). For non-normally distributed data, comparisons between multiple groups were performed using the Kruskal-Wallis test, followed by Dunn’s multiple comparisons test (GraphPad Prism 9.0). Differences were considered statistically significant at p 0.05, p 0.01, and p 0.001. A list of key resources used in this study was shown in Supplementary Table 1.

CatWalk gait analysis was performed to objectively evaluate pain-induced locomotor alterations. This method has been used previously to assess mechanical allodynia in animal models [17]. As shown by the timing view and footprint intensities (Fig. 1A), CFA-treated mice exhibited significant alterations in their bilateral hindlimb stepping pattern. Compared with baseline measurements, CFA-treated mice showed notable reductions in stand phase duration (Fig. 1B), duty cycle (Fig. 1D), single stance duration (Fig. 1E), and max contact max intensity (Fig. 1F) of the affected hindpaw, together with an increase in the swing phase duration (Fig. 1C). These abnormalities were most prominent during the first 3 days post CFA injection, suggesting peak spontaneous pain during this period. However, gait parameters showed a gradual recovery beginning on day 5. To exclude potential motor dysfunction induced by CFA injection, we also performed Rotarod testing. Both the latency and maximal rotational velocity at fall remained unchanged in CFA-treated mice compared with controls (Fig. 1G), indicating the observed gait abnormalities were primarily pain-driven rather than because of motor impairment. We also examined complementary weight-bearing balance and reflexive pain-related behaviors. Both these measurements revealed that CFA-treated mice exhibited sustained spontaneous pain and evoked hypersensitivity, especially in von Frey filament-evoked mechanical allodynia (Fig. 1H,I). While gait analysis indicated partial recovery by at least day 5, weight-bearing asymmetry and decreased mechanical withdrawal thresholds persisted through day 7. Collectively, these behavioral analyses demonstrate that intraplantar CFA injection induces both dynamic temporal changes and persistent static pain.

Fig. 1.

CFA-induced inflammatory pain induces time-dependent changes in spontaneous and evoked pain behaviors. (A) Representative CatWalk gait analysis showing timing views (left) and pressure intensity maps (right) of hindpaw prints in control mice and at the indicated time points after CFA injection (days 1, 3, 5 and 7). (B–F) Quantitative analysis of various gait parameters assessed by CatWalk: (B) stand phase duration, (C) swing phase duration, (D) duty cycle, (E) single stance duration, and (F) max contact max intensity. Data were collected from control mice and CFA-treated mice. Values were normalized as the ipsilateral (affected left hindpaw)/contralateral (unaffected right hindpaw) ratio. Mean $\pm$ SEM (control, n = 8, CFA-treated, n = 9; * p 0.05, ** p 0.01, *** p 0.001 vs. control; two-way ANOVA with Šídák’s test). (G) Motor coordination was assessed by the Rotarod test on day 1 post-CFA injection. Left: latency to fall (seconds); Right: maximal rotational velocity at fall (rpm). Mean $\pm$ SEM (control, n = 8, CFA-treated, n = 9; unpaired Student’s t-test). (H) Weight bearing measured as the difference between ipsilateral and contralateral hindpaws (grams) after CFA injection. Mean $\pm$ SEM (Control, n = 8; CFA-treated, n = 8; ** p 0.01, *** p 0.001 vs. control; two-way ANOVA with Šídák’s test). (I) Mechanical hypersensitivity measured by the von Frey filament test and showing paw withdrawal thresholds for the ipsilateral hindpaw. Mean $\pm$ SEM (control, n = 8, CFA-treated, n = 8; ** p 0.01, *** p 0.001 vs. control; two-way ANOVA with Šídák’s test). Adobe Illustrator 2020 (version 24.1.3, Adobe Inc., San Jose, CA, USA) was used to create the illustrations in (G–I). CFA, complete Freund’s adjuvant; SEM, standard error of the mean; ANOVA, analysis of variance; BL, baseline; LH, left hindpaw; RH, right hindpaw.

Protein expression of the immediate early gene c-Fos has been widely used as a marker for neuronal activation [18]. To study dynamic and laminar-specific sensory processing in the SDH, we examined the temporal and spatial distribution of c-Fos expression across distinct SDH laminae following CFA injection. IB4 and PKC$\gamma$ were employed as laminar-specific markers to delineate the inner dorsal (II_id) and ventral (II_iv) boundaries of lamina II, respectively [19] (Fig. 2A). SDH regions were categorized accordingly into superficial (laminae I–II) and deep (laminae III–V) domains. Quantitative analysis revealed minimal c-Fos expression in the superficial laminae of control mice. As expected, CFA induced a significant increase in c-Fos positive neurons within the superficial laminae on day 1, when the mice displayed obvious nociceptive behaviors (Fig. 2B,D). However, the elevated c-Fos expression decreased to normal levels after day 3. In addition, the number of c-Fos positive neurons in the deep laminae remained comparable with control levels (Fig. 2C,D). These results suggest that enhanced neuronal activation in superficial SDH laminae may represent a key neural substrate underlying the manifestation of spontaneous pain in CFA-induced inflammation.

Fig. 2.

Time-dependent c-Fos expression across superficial and deep laminae of the SDH after CFA injection. (A) Schematic representation of mouse spinal dorsal horn showing lamina II_i delineated by IB4 (green) and PKC$\gamma$ (white) immunostaining. (B,C) Quantitative analysis of c-Fos positive neurons in superficial (lamina I–II, B) and deep laminae (lamina III–V, C) of the dorsal horn at indicated time points after CFA injection. (n = 4 mice per group). Mean $\pm$ SEM (* p 0.05 vs. control; Kruskal-Wallis test with Dunn’s test). Effect sizes and 95% confidence intervals of the data was shown in Supplementary Table 2. (D) Representative immunofluorescence images showing c-Fos expression (red) in superficial (lamina I–II) and deep laminae of the dorsal horn in control and CFA-treated mice across different time points (days 1, 3, 5, and 7). Scale bars: 50 µm. Adobe Illustrator 2020 (version 24.1.3, Adobe Inc.) was used to create the illustrations in (A). SDH, spinal norsal horn; IB4, isolectin B4; PKC$\gamma$, protein kinase C-gamma.

A comprehensive view of real-time neuronal responsiveness in persistent inflammatory pain was obtained using multi-lamina extracellular recordings to characterize the spatiotemporal dynamics of neuronal excitability throughout the dorsal horn. Conventional c-Fos mapping fails to detect this. Neurons in the SDH were categorized into superficial (0–250 µm) and deep regions (250–850 µm) based on the locations of their recording (Fig. 3A–C, Ref. [20] for C). Each experiment yielded an average of 12 isolated single units per mouse.

Fig. 3.

Temporal changes in spontaneous and evoked neuronal activity across different depths of the SDH after CFA injection. (A) Experimental setup for in vivo SDH neuronal recordings. A 32-channel MEA was used to record action potentials from SDH neurons while applying mechanical stimuli (touch, brush and pinch) to the ipsilateral hindpaw receptive fields. (B) Configuration of the 32-channel microelectrode array showing the recording sites distributed across an 850 µm depth. (C) Predicted extent and borders of spinal laminae across the 850 µm depth for 300 µm lateral to the middle spinal artery. The SDH laminar organization were divided into superficial (0–250 µm) and deep (250–850 µm) regions. The annotated L4 spinal cord section was adapted from the Mouse Spinal Cord Atlas provided by the Allen Institute for Brain Science (https://mouse.brain-map.org/experiment/siv?id=100050402&imageId=101006501&imageType=atlas). (D–G) Depth-dependent neuronal activity represented as heatmaps and showing (D) spontaneous, (E) brush-evoked, (F) touch-evoked, and (G) pinch-evoked firing rate changes (Hz). Depth bin for each row was 25 µm. The color scale represents the firing frequency in Hz. Animal numbers: control: n = 6; day 1: n = 6; day 3: n = 6; day 5: n = 7; day 7: n = 6. Temporal profile of spontaneous and evoked neuronal activity in the superficial and deep dorsal horn following CFA injection. A detailed analysis of spontaneous and evoked neuronal activity across the superficial and deep SDH is presented in Supplementary Fig. 5 and Supplementary Table 3. Adobe Illustrator 2020 (version 24.1.3, Adobe Inc.) was used to create the illustrations in (A–C). MEA, microelectrode array.

Heatmap analysis showed laminar-specific differences in spontaneous and evoked neuronal activity between control and CFA-treated mice. In control animals, neurons in the superficial laminae exhibited minimal activity, whereas deep laminae displayed higher levels of spontaneous firing (Fig. 3D). Following CFA injection, the spontaneous excitabilities of neurons among all laminae increased rapidly on day 1, then gradually declined to normal levels (Fig. 3D). As for evoked activity, neurons across all laminae displayed limited responses to mechanical stimuli under physiological conditions. In mice injected with CFA, brush stimulus elicited the most pronounced neuronal activity in the superficial laminae, whereas touch and pinch stimuli induced comparable neuronal hyperactivity across all laminae (Fig. 3E–G).

Collectively, these electrophysiological recordings suggest that neurons across the dorsal horn in mice with inflammatory pain tend to exhibit an increase in spontaneous and evoked excitability. The temporal profile of these alterations correlates well with both spontaneous and evoked pain behaviors, providing a more comprehensive understanding of the neuronal basis of inflammatory pain processing.

To further elucidate the neuronal population underlying inflammatory pain, the proportion and distribution of SDH neurons were characterized across three functional subtypes: WDR, LTM, and NS neurons. Mechanical response was assessed using dynamic brush and static touch as innocuous stimulation, and a 150-g calibrated forceps pinch was used as noxious stimulation. Neurons responding to both stimuli were classified as WDR, while neurons responding exclusively to innocuous stimuli were LTM, and neurons responding exclusively to noxious stimuli were NS (Fig. 4A). Quantitative analysis revealed dynamic shifts in the composition of mechanically-responsive neuronal populations (Fig. 4B). Under physiological conditions, WDR neurons constituted the predominant population (44.44%) of recorded neurons, whereas LTM and NS neurons were present in comparable proportions (27.78%). On day 1 after CFA injection, the proportion of WDR neurons decreased markedly (36.99%) but returned to prominence in later observations. LTM and NS neurons exhibited divergent dynamics at day 1, with LTM peaking at 41.10%, whereas NS neurons declined to 21.92%. Both neuronal subtypes gradually returned to baseline levels thereafter.

Fig. 4.

Temporal changes in neuronal subtypes and their spatial distribution in SDH following CFA-induced inflammatory pain. (A) Representative response patterns defining neuronal subtypes. WDR neurons show graded responses to brush, touch, and pinch stimuli. LTM neurons respond only to brush stimulation, while NS neurons respond exclusively to pinch stimulation. (B) Proportion of neuronal subtypes (WDR, LTM, and NS) in control mice and inflammatory pain mice at indicated time points. Values are shown as the percentages of total recorded neurons. (C–E) Spatial distribution analysis of recorded neurons across dorsal horn depth: (C) WDR neurons (control, N = 16; day 1, N = 27; day 3, N = 40; day 5, N = 32; day 7, N = 27). (D) LTM neurons (control, N = 10; day 1, N = 30; day 3, N = 28; day 5, N = 27; day 7, N = 12). (E) NS neurons (control, N = 10; day 1, N = 16; day 3, N = 17; day 5, N = 17; day 7, N = 16). N represents neuron number (animal n values match Fig. 3). Scatter plots display the depth distribution, where light, and dark gray regions correspond to superficial and deep laminae, respectively. Values are presented as the mean $\pm$ SEM. Statistical significance was determined using Kruskal-Wallis test followed by Dunn’s multiple comparisons. * p 0.05, ** p 0.01 vs. control. WDR, wide dynamic range; LTM, low-threshold mechanoreceptive; NS, Nociceptive-specific.

Spatial distribution analysis revealed a shift in depth of the neuronal populations following inflammatory pain. WDR neurons, which normally predominate in deeper laminae, exhibited a significant shift toward more superficial locations on days 3 and 5 (Fig. 4C, Supplementary Table 4). LTM neurons were typically concentrated in deep laminae under normal conditions, but showed an enhanced superficial distribution on days 1, 3, and 5 post-CFA compared with controls (Fig. 4D, Supplementary Table 4). The distribution of NS neurons remained unchanged throughout the observation period (Fig. 4E, Supplementary Table 4). The concurrent superficial shift of both LTM and WDR neurons suggests enhanced nociceptive processing through increased integration of both noxious and innocuous inputs in pain-processing circuits. Overall, alterations in both the proportion and distribution of SDH neurons may represent a form of maladaptive plasticity that contributes to the heightened pain sensitivity observed in our behavioral studies.

After characterizing the changes in proportion and distribution of SDH neurons, we next assessed their mechanical excitabilities by quantifying evoked discharges in response to various stimuli (brush, touch, and pinch) following CFA injection. The discharge frequency of WDR neurons evoked by brush and pinch stimuli was significantly higher on days 1 and 3 post CFA injection compared with controls. Additionally, these neurons showed increased mean firing frequency to touch specifically on day 1 (Fig. 5A–C). Interestingly, despite their altered proportion and distribution patterns, LTM neurons in mice with inflammatory pain showed a slight increase in their brush-evoked excitabilities (Fig. 5D). NS neurons maintained relatively stable activities to pinch stimuli throughout the observation period (Fig. 5E). The differential evoked activities of WDR, LTM, and NS neurons suggests that WDR neurons serve as key cellular mediators for the manifestation of both mechanical allodynia and hyperalgesia in inflammation.

Fig. 5.

Temporal changes in stimulus-evoked responses of distinct neuronal subtypes after CFA injection. (A–C) Analysis of WDR neuron firing rates in response to (A) brush, (B) touch, and (C) pinch stimulation. (D) Brush-evoked firing rates of LTM neurons. (E) Pinch-evoked firing rates of NS neurons. Values are presented as individual data points on violin plots (N values for neurons as shown in Fig. 4). Statistical significance was determined using Kruskal-Wallis test followed by Dunn’s multiple comparisons. * p 0.05 vs. control. See Supplementary Fig. 6 for a detailed analysis of WDR neuron responses to mechanical stimuli across days.