Therapeutic potential of 5HT2CR in alleviating spasticity and pain via modulation of motor neurons and interneurons in spinal cord injury

Summary

Numerous adverse outcomes can be resulted in limb spasm and spasticity that occur after spinal cord injury (SCI). Here, we used adeno-associated virus (AAV) vectors and Cre-driver transgenic mice to selectively overexpress 5-hydroxytryptamine2C receptor (5HT2CR) in motor, GABAergic, and Glutamatergic interneurons. Targeted overexpression in these neuronal populations attenuated spasticity-like responses and hindlimb motor deficits, and improved sensory abnormalities and affective disturbances. These behavioral improvements were associated with changes in potassium chloride cotransporter-2 (KCC2), gamma-aminobutyric acid type A receptor (GABAAR), and N-methyl-D-aspartate receptor (NMDAR) expression in spinal cord tissue, indicating that the activity of motor neurons is influenced by the expression of 5HT2CR. This work indicates that 5HT2CR represents a promising target for alleviating spasticity and neuropathic pain after SCI.

Graphical abstract

Highlights

• •Overexpression of 5HT2CR on MNs or INs alleviates the degree of spasticity after SCI
• •Overexpression of 5HT2CR on MNs improves motor function and reduces anxiety/depression
• •Overexpression of 5HT2CR on Glu INs better alleviates neuropathic pain
• •Overexpression of 5HT2CR modulates the expression levels of KCC2, GABAAR, and NMDAR

Biological sciences; Natural sciences; Neuroscience; Sensory neuroscience

Introduction

Muscle spasticity is a common complication of spinal cord injury (SCI), representing a sensory-motor control disorder following upper motor neuron damage, characterized by intermittent or persistent involuntary muscle activation.¹ Approximately two-thirds of SCI patients experience varying degrees of muscle spasticity within the first six months after injury. Spasticity can result in stiffness of the limbs, sleep disturbances, spasmodic pain, and even spastic paralysis, significantly impacting patients’ daily care, rehabilitation training, and quality of life.^2,3,4 Treatment for spasticity after SCI encompasses both conservative and surgical interventions.^5,6,7 While the current therapeutic modalities lack curative efficacy and exhibit inherent deficiencies,⁸ it is imperative to seek effective relief and treatment methods for spasticity after SCI.

The integrity of the 5HT (5-hydroxytryptamine or serotonin) system in the spinal cord is crucial for regulating sensation and coordinating motor functions.⁹ Cental serotonin deficiency impairs sensory-motor recovery after SCI.¹⁰ The severity of spinal cord injury correlates positively with the extent of damage to 5HT neuron axons, making the 5HT system a promising target for treating muscle spasticity post-injury.^11,12 Various 5HT receptors, such as serotonin 1A,^13,14 1B,¹⁵ 2A,¹⁶ 2B,¹⁷ and 2C,¹⁸ have been investigated for their therapeutic potential in addressing SCI and its complications. Following SCI, disruption of 5HT projections results in reduced serotonin levels at the injury site.¹⁹ This depletion may lead to hypersensitivity or constitutive activation of specific 5HT receptors involved in spinal cord network regulation. To date, numerous strategies targeting the 5HT system have shown potential in improving motor function and coordination in rodent models of spinal cord injury. The 5HT system directly affects neuronal excitability, which depends on the location and activation of 5HT receptors that modulate Na⁺, K⁺, and Ca²⁺ channel states.²⁰ Furthermore, the 5HT system regulates glutamatergic and GABAergic neurotransmissions, indirectly influencing motor neuron excitability by acting on spinal interneurons.²¹ Thus, enhancing the 5HT system’s function may alleviate sensory and motor deficits, reduce muscle spasms, and help restore normal physiological functions following spinal cord injury.

The mechanisms of spasticity after spinal cord injury are complex and still debated. These mechanisms includes: (1) hyperactivity of fusiform muscle movement; (2) reduction of post-activation depression (PAD); (3) reduction of presynaptic inhibition; (4) reduction of Ia reciprocal inhibition; (5) increased excitability of motor neurons; (6) sustained activation of persistent inward currents (PICs); (7) differential regulation of spinal cord neurons by descending drive from monoamine neurotransmitters; (8) dysregulation of excitation/inhibition (E/I) input in interneurons.⁵ However, the direct cause of spasticity is the excessive excitability of motor neurons. Therefore, it is an urgent challenge to balance excitatory and inhibitory signals, regulate motor neuron activity, and suppress the occurrence and progression of limb spasticity, all while not hindering the recovery of motor function after spinal cord injury.

The regulation of the descending monoaminergic system may be a crucial breakthrough for achieving the aforementioned goals. Part of the cause of muscle spasticity following spinal cord injury is associated with the loss of the brainstem-derived serotonergic system. It has been found that 5HT2CR can regulate motor neuron excitability by modulating persistent inward currents (PICs). Although this modulation can enhance muscle contraction function lost after SCI, it also contributes to increased spasticity.²²

In our previous studies, a finding that piqued our interest, we used adeno-associated virus (AAV) to overexpress neurotrophin-3 (NT-3) and insulin-like growth factor-1 (IGF-1) in SCI rats, which alleviated hindlimb spasticity and showed an increasing trend in 5HT2CR expression.²³ Therefore, in this study, we evaluated the effects of overexpressing 5HT2CR in motor neurons, inhibitory interneurons, and excitatory interneurons. A transgenic mouse model with hindlimb spasticity following SCI, which carries the Cre enzyme gene, was used. By utilizing AAV viruses with DIO elements, 5HT2CR was selectively overexpressed in different types of neurons. The aim of this study was to evaluate the differential impact of 5HT2CR overexpression in these three types of neurons on motor function and spasticity symptoms in SCI mice.

We demonstrated that 5HT2CR overexpression can alleviate spasticity in an SCI mouse model. By using transgenic mice, we successfully induced 5HT2CR overexpression in various types of spinal cord neurons. Our findings show that overexpressing 5HT2CR in motor neurons significantly improves muscle spasticity, motor function, and gait during the chronic phase of SCI, while overexpressing 5HT2CR in interneurons better alleviates neuropathic pain.

Results

AAV-mediated 5HT2CR gene overexpression in target neurons: Focus on MNs and INs

This experiment is based on the theory of the Cre-LoxP system. Firstly, we engineered an AAV, designated HBAAV2/BBB-CAG-DIO-m-Htr2c-3xflag-ZsGreen, with the capability to cross the blood-brain barrier. This AAV vector incorporates both the 5HT2CR gene fragment and a green fluorescent marker (Figures 1A and 1B).

Figure 1.

Selective overexpression of 5HT2CR in spinal neuronal populations and its impact on H-reflex modulation

(A) Experimental groups.

(B) Illustration of viral constructure and the genetic cross strategy using ChAT-IRES-Cre, VGAT- IRES-Cre, and Vglut2- IRES-Cre mice.

(C) Representative confocal images showed motoneurons、GABAergic interneurons or glutamatergic interneurons transfected with AAV in lumbar enlargement of spinal cord.

(D) Diagram of experimental design.

(E) Principle of the H-reflex.

(F) Classic electrophysiological trace.

(G) RDD of the H -reflex at various stimulus frequencies from 0.2Hz to 5Hz, comparing SCI, VGAT-Cre, CHAT-Cre, and Vglut-2-Cre groups at different time points (35dpi, 56dpi).n = 5 in each group. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001 for VGAT vs. SCI, ^$p < 0.05, ^$$p < 0.01, ^$$$p < 0.001 for ChAT vs. SCI; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 for VGluT2 vs. SCI. Data are presented as means ± SEM.

To evaluate the effects of 5HT2CR on spasticity, motor function, and sensory processing following SCI, we divided mice into five groups: the sham surgery group (laminectomy without SCI), the SCI group, the Chat group (SCI with AAV-mediated overexpression of 5HT2CR in motor neurons), the VGAT group (SCI with AAV-mediated overexpression of 5HT2CR in GABAergic interneurons), and the Vglut-2 group (SCI with AAV-mediated overexpression of 5HT2CR in glutamatergic interneurons). Mice were pre-trained for seven days prior to the induction of the SCI surgery. Following laminectomy at the T9 vertebral level, we generated the SCI model by delivering a vertical impact injury to the T10 spinal cord using a drop-weight device. On the second day post-injury (dpi), the AAV encoding 5HT2CR was diluted to 1.5 × 10ˆ10 vg/mL (100-fold dilution) using sterile 0.9% saline. This solution was then administered via tail vein injections into ChAT-IRES-Cre, VGAT-IRES-Cre, and Vglut2-IRES-Cre mice (Figures 1A and 1B).

Immunofluorescence was subsequently utilized to co-label ChAT, VGAT, and Vglut-2, demonstrating successful AAV transduction in the target cells. This was evidenced by the examination of transverse sections of the lumbar enlargement spinal tissue, where robust green fluorescent labeling was consistently observed across all experimental groups. At 28 dpi, one mouse from each group was randomly selected for detailed immunofluorescence analysis. Quantitative co-localization analysis revealed that most of the green fluorescent markers were localized in the expected target cells. These results provide strong evidence of targeted and efficient gene delivery to ChAT, VGAT, and Vglut-2 expressing cells, validating the specificity and effectiveness of the AAV-mediated transduction (Figure 1C).

H-reflex tests were conducted at weekly intervals to monitor changes in spasticity in the mice on 35 dpi and 56 dpi. Additionally, the locomotor function of the mice was evaluated on both 28 dpi and 56 dpi using the Basso mouse scale (BMS) to provide quantitative scoring. On 56 dpi, a comprehensive assessment was performed, including Digit Gait Analysis for detailed gait metrics, pain threshold measurements, and the open field test (OFT) to evaluate anxiety and depression-related behaviors. These evaluations collectively provided a thorough assessment of the neurological and functional recovery in the mice (Figure 1D).

The effect of 5HT2CR overexpression in MNs and INs on spasticity

To evaluate spastic-like behavior, we measured the Hoffman reflex (H-reflex) in the hind limbs of mice. The amplitude of the H-reflex decreases with increasing stimulation frequency under continuous stimulation, a phenomenon known as rate-dependent depression (RDD). We utilized the RDD of the H-reflex to assess the extent of this suppression.²⁴ To investigate the changes in spastic-like symptoms in the hind limbs of these SCI mice, RDD of the H-reflex were measured at 35 dpi and 56 dpi.

We recorded both the M-wave and the H-wave. The M-wave, which is generated by the activation of motor axons, showed minimal variation among the different groups of mice. While the Ia sensory fibers mediate the excitation of motor neurons upon stimulation, resulting in a delayed H-wave (Figures 1E–1G). As the stimulation frequency increases, the amplitude of the H-wave diminishes. However, in injured mice exhibiting muscle spasms due to excessive excitation of motor neurons, the reduction in H-wave amplitude is suppressed.²⁵ Consequently, as the stimulation frequency increases from 0.1 Hz to 5 Hz, animals with mild spasticity show a significant decrease in RDD, whereas animals with severe spasticity exhibit the opposite pattern.

In this experiment, the average relative amplitude of the H-wave (H%) in mice decreased when the stimulation frequency reached or exceeded 0.5 Hz. The RDD of H-wave in the SCI group was significantly higher than injured mice in the other three groups which overexpressed 5HT2CR. At 35 dpi, the decrease in H-wave amplitude in the SCI group was not significant (RDD decreased from 100% at 0.2 Hz to 89.73 ± 25.99% at 5 Hz). In contrast, the H% in the VGAT, ChAT, and Vglut-2 groups was significantly lower than that in the SCI group (decreasing to 36.70% ± 26.20%, 58.59% ± 16.5%, and 46.93% ± 24.90% at 5 Hz, respectively, all p < 0.05 compared to SCI group) (Figure 1G). Similar RDD patterns were observed at 56 dpi, with the SCI group showing 81.8% ± 21.80% at 5 Hz, compared to 22.80% ± 21.08% in the VGAT group, 23.80% ± 15.62% in the ChAT group and 50.2% ± 10.77% in the Vglut-2 group (Figure 1G).

These findings indicate that overexpressing 5HT2CR in both motor neurons and interneurons can significantly reduce the excessive excitation of motor neurons. This, in turn, effectively alleviates spastic symptoms observed in the hind limbs following SCI.

5HT2CR overexpression in motor neurons reduces anxiety-depressive behaviors in SCI mice

After SCI, anxiety and depressive-like behaviors are commonly observed, and these negative emotions may be exacerbated by muscle spasticity and pathological neuropathic pain. Serotonin, an important target for treating anxiety and depression, has rapid antidepressant effects. The OFT is an effective method for measuring locomotor activity and anxiety-like behavior.²⁶ Exploration of the center area is commonly considered an indicator of the anxiety and depression levels in mice. To investigate the impact of overexpression of 5HT2CR in different types of neurons on anxiety behavior following SCI, the OFT was used to observe the effects of 5HT2CR overexpression in motor neurons or interneurons on the emotional state of spinal cord injured mice. The movement trajectories of the mice in the open field were recorded (Figure 2C). The speed and distance traveled by mice through the all area and the center area were evaluated. Meanwhile, attention was paid to the number of times mice passed through the center area and the duration of their traversal in the center area (Figures 2A and 2B).

Figure 2.

Behavioral analysis of mice in different experimental groups

(A and B) Schematic of the open field test setup used to evaluate anxiety-like behavior.

(C) Representative tracking paths taken by mice within the open field, shown for each experimental group.

(D–G) Distance traveled, velocity, total movement, and central zone time were assessed in different experimental groups. n = 3 in each group. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001. ns indicates no statistical difference between groups.

(H) Schematic of the Digi Gait system analysis setup used to evaluate locomotor activity.

(I)Top: Right hindlimb stride time, braking time, stance time, and stride frequency, hindlimb stride frequency; hindlimb stance width. Bottom: Left hindlimb stride time, braking time, stance time, and stride frequency, hindlimb stride frequency; hindlimb stance angle. n = 5 in each group. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001.ns indicates no statistical difference between groups.

(J) Schematic of the BMS used to evaluate locomotor activity.

(K) Top: BMS at 35dpi.Bottom: BMS at 56dpi.n = 5 in each group. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001 for ChAT, VGAT,Vglut-2 group vs. SCI group;^#p < 0.05, ^##p < 0.01, ^###p < 0.001 for ChAT, VGAT, Vglut-2 group vs. Sham group. ns indicates no statistical difference between groups. Data are presented as means ± SEM.

The traversal velocity and distance across all areas for mice of ChAT, VGAT, and Vglut-2 groups showed no significant improvement compared to those in the SCI group (p > 0.05). Specifically, the Vglut-2 group demonstrated minimal variation across any observed metrics relative to the SCI group, with a notable reduction in velocity through the central area (15.53 cm/s ± 3.82 cm/s, p < 0.001). Contrastingly, during traversal of the central area, ChAT group mice exhibited a significantly greater velocity compared to the SCI group (p < 0.05). No significant difference in speed was observed between the ChAT and Sham group (p > 0.05). Mice in the ChAT group traveled 175.31 mm ± 17.62 mm, which was significantly greater than the SCI group (83.83 mm ± 6.50 mm,p < 0.001) but notably less than the Sham group (911.16 mm ± 164.55 mm,p < 0.05) (Figures 2D–2E). The duration time in the center area of the open field for mice in the ChAT group was significantly higher than that for mice in the SCI group (p = 0.0304) (Figure 2F) The ChAT group did not show any significant difference in bouts in the center compared to both the Sham groups (p < 0.05) (Figure 2G).

In summary, the ChAT group mice exhibited better exploratory ability. An effect that was less pronounced in the VGAT and Vglut-2 groups. These results indicate that overexpression of 5HT2CR in motor neurons alleviates anxiety and depressive-like symptoms in SCI mice. However, overexpression of 5HT2CR in interneurons does not improve anxiety and depressive-like behaviors.

Modulation of motor function and gait dynamics through 5HT2CR overexpression

SCI results in motor impairments, primarily due to the loss of descending signals from higher central centers following spinal cord transection. Additionally, spasticity further exacerbates these impairments, specifically affecting movement and gait. We performed a multidimensional evaluation of hindlimb motor function in mice by integrating gait analysis with the DigiGait system and BMS scoring.^27,28

Utilizing the DigiGait gait analysis system, we examined the impact of 5HT2CR overexpression in distinct neuronal populations on mouse gait patterns (Figure 2H). Key gait parameters, such as stride time, braking time, stance time, stride frequency, hindlimb stance width, and hindlimb step angle, were rigorously analyzed. Due to poor recovery of motor function after injury, the SCI group mice were unable to complete the gait recording process using the DigiGait system. The results showed that compared to the control group, the Vglut-2 group exhibited significant differences in gait parameters compared to the Sham group, particularly in stance time and braking time. The stance time was significantly prolonged in the Vglut-2 group (p < 0.05 for the right hindlimb, p < 0.01 for the left hindlimb). Moreover, braking time was significantly extended in both hind limbs, with a particularly pronounced increase in the left hindlimb (p < 0.001), indicating suboptimal regulatory effects in the Vglut-2 group (Figure 2I).

In contrast, the ChAT and VGAT groups demonstrated notable improvements across various gait parameters, with no significant abnormalities observed in stance time or stride frequency (Figure 2I). This suggests enhanced motor autonomy and coordination. In addition, the Vglut-2 group showed a significant increase in stance width and hindlimb stance angles (p < 0.05), but this did not result in positive gait adjustments. While the ChAT and VGAT groups demonstrated stable performance in these parameters.

Overall, the overexpression of 5HT2CR did not effectively improve gait characteristics in the Vglut-2 group, whereas the ChAT and VGAT groups demonstrated superior regulatory effects.

At 35 dpi and 56 dpi, motor function was assessed using the BMS scoring system. (Figure 2J). The results revealed no significant difference in BMS scores between the Vglut-2 group and the SCI group (p > 0.05), aligning with the prolonged gait cycle and braking times observed in the gait analysis. This finding further underscores the limited impact of Vglut-2 in modulating gait function. In contrast, mice in the ChAT and VGAT groups exhibited significant improvements in BMS scores, highlighting their superior efficacy in enhancing motor recovery (p < 0.01 and p < 0.05, respectively) (Figure 2K).

This outcome is consistent with the marked enhancements observed in gait analysis, indicating that elevating the expression of 5HT2CR in motor neurons or GABAergic interneurons can offer significant potential for recovering gait function and alleviating motor impairments after SCI.

Overexpression of 5HT2CR in Glu interneurons modulates mechanical and thermal pain thresholds after SCI

Hyperalgesia and abnormal pain can be quantitatively assessed in different mouse groups through thermal and mechanical pain thresholds using the Hargreaves and Von Frey tests, respectively. At 55 dpi, thermal pain thresholds were determined across various mouse cohorts. The Vglut-2 group exhibited significantly elevated thermal thresholds relative to the SCI group (p < 0.01). No significant differences in thermal pain thresholds were observed between the ChAT and VGAT groups compared to the SCI group (Figure 3A). To reduce potential confounding factors, the Von Frey test was conducted the following day. The results revealed that mechanical pain thresholds in the ChAT, VGAT, and Vglut-2 groups were markedly higher than those in the SCI group (p < 0.0001, p < 0.05, p < 0.001, respectively) (Figure 3B). These data underscores that overexpression of 5HT2CR in interneurons more effectively ameliorates pain hypersensitivity in SCI mice.

Figure 3.

Pain Assessment and Expression of NMDAR1 and NMDAR2A in spinal cord tissue

(A) Left: representative schematic of the Hargreaves test used to measure the thermal pain threshold. Right: latency to thermal pain in different groups. n = 6 in each group.∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001 for SCI, ChAT, VGAT, Vglut-2 group vs. SCI,^#p < 0.05, ^##p < 0.01, ^###p < 0.001 for ChAT, VGAT, Vglut-2 group vs. Sham. n = 6 in each group. Data are presented as means ± SEM.

(B) Left: Representative schematic of the Von Frey test used to measure the mechanical pain threshold. Right: Paw withdrawal threshold to mechanical pain in different groups. n = 6 in each group.∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001 for SCI, ChAT, VGAT, Vglut-2 group vs. SCI,^#p < 0.05, ^##p < 0.01, ^###p < 0.001 for ChAT, VGAT, Vglut-2 group vs. Sham. n = 6 in each group. Data are presented as means ± SEM.

(C–D) Immunofluorescence staining for NMDAR1 and NMDAR2A punctae on lumbar motor neurons in the SCI group, ChAT group, VGAT group, and VGLUT-2 group.n = 3 mice in each group at 8 weeks. The inserts were magnified to show the anterior horn of spinal cord. Scale bars: 50 mm.

(E) Quantification of NMDAR2 and NMDAR2A positive punctae on lumbar motor neurons shown in (C–D). n = 3 mice in each group at 8 weeks. Data are presented as means ± SEM. ∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001. Data are presented as means ± SEM.

Overexpression of 5HT2CR in MNs or INs regulates the expression levels of GABAAR and KCC2 in motor neurons

Immunofluorescence analysis was conducted to assess the expression levels of NMDAR1 and NMDAR2A in the lumbar enlargement of different mouse groups. In the SCI group, the expression levels of both NMDAR1 and NMDAR2A were elevated compared to the Sham group (NMDAR1,210.25 ± 36.60 for the SCI group, 158.5 ± 23.10 for the Sham group, p > 0.05; NMDAR2A, p < 0.001) (Figures 3C–3E). This indicates that SCI induces an abnormal upregulation of NMDAR expression. In the ChAT and VGAT groups, the number of NMDAR1-positive cells did not differ significantly from the Sham group, whereas the Vglut-2 group exhibited a notable increase in NMDAR1 expression (p < 0.05). Additionally, the number of NMDAR2A-positive cells in the lumbar enlargement of ChAT and VGAT groups was significantly lower than that in the SCI group (p < 0.05,p < 0.05, respectively, but no significant difference was observed between the Vglut-2 and SCI groups (p > 0.05) (Figures 3C–3E). These findings suggest differential regulation of NMDAR subunits in response to SCI and its modulation by various neuronal subtypes.

In summary, overexpression of 5HT2CR in motor neurons and GABAergic interneurons of SCI mice significantly reduced the expression of NMDAR1 and NMDAR2A in the lumbar enlargement. In contrast, overexpression of 5HT2CR in glutamatergic interneurons had no significant effect on NMDAR expression in SCI mice.

The effects of 5HT2CR overexpression in different neurons on GABAAR and KCC2

Initially, at 56 dpi, we conducted a Western Blot analysis on the entire spinal cord of mice to measure the expression levels of BDNF, TrkB, KCC2, and GABAAR. The results indicated no significant differences in the expression of BDNF and TrkB between the VGAT, ChAT, Vglut-2, and SCI groups (p > 0.05) (Figures 4A and 4B). However, the expression levels of GABAAR in the VGAT and ChAT groups were significantly lower compared to the SCI group (p = 0.0025 and p = 0.0001, respectively). Additionally, the expression of KCC2 was significantly reduced in the VGAT and Vglut-2 groups compared to the SCI group (p = 0.021 and p = 0.005, respectively) (Figures 4A and 4B). Therefore, the overexpression of 5HT2CR does not modulate the BDNF-TrkB pathway.

Figure 4.

Expression and localization of neurotransmitter receptors and transporters in spinal cord tissue

(A and B) Immunofluorescence staining for KCC2 and GABAAR punctae, as well as co-labeling of ChAT, showing the immunofluorescence intensity on lumbar motor neurons in the SCI group, ChAT group, VGAT group, and VGLUT-2 group.n = 4 mice in each group at 8 weeks. The inserts were magnified to show the anterior horn of spinal cord. Scale bars: 50 mm.

(C and D) Quantification of immunofluorescence intensity for GABAAR and KCC2 on MN shown in (A) and (B). n = 3 mice in each group at 56pdi. Data are presented as means ± SEM. ns indicates no statistical difference between groups.

(E) Representative western blot of BDNF, TrkB, KCC2,and GABAAR of spinal cord tissue, showing the expression levels of those protein in the SCI group, Sham group, ChAT group, VGAT group, and VGLUT-2 group. GAPDH was used as a loading control. n = 3 mice in each group at 8 weeks.

(F) Quantification of those protein levels between the four groups, normalized to GAPDH. n = 3 mice in each group at 8 weeks. Data are presented as means ± SEM.∗p < 0.05, ∗∗p < 0.01,∗∗∗p < 0.001. Data are presented as means ± SEM.

Next, we focused on the lumbar enlargement region of the mouse spinal cord, assessing the expression of GABAAR and KCC2 in motor neurons via immunofluorescence. The mean intensity of these proteins in motor neurons was our primary metric of interest (Figures 4C and 4D). The results indicated that there were no significant differences in the expression levels of GABAAR and KCC2 between the VGAT, ChAT, and Vglut-2 groups when compared to the SCI group (Figures 4E and 4F). However, both the VGAT and ChAT groups exhibited a trend toward decreased mean intensity of GABAAR and increased mean intensity of KCC2 in motor neurons relative to the SCI group. Conversely, the Vglut-2 group showed an increasing trend in the mean intensity of GABAAR and a decreasing trend in the mean intensity of KCC2 compared to the SCI group (Figures 4E and 4F). In summary, the overexpression of 5HT2CR regulates the expression and function of GABAAR and KCC2 in the spinal cord.

Discussion

Serotonin reuptake inhibitors, a class of antidepressants, have been reported to exacerbate spasticity associated with SCI as early as the last century.^29,30 5HT2CR has been identified in various neuronal populations. The Cre-LoxP system has been utilized to successfully achieve overexpression of 5HT2CR in motor neurons, GABAergic interneurons, and glutamatergic interneurons in SCI mice. These neurons demonstrate the physiological characteristics of spinal motor circuits in vitro. The expression of 5HT2CR, which eventually regulates the function and excitability of motoneurons in the lumbar enlargement of the spinal cord, influences spastic-like symptoms, motor function, neuropathic pain, and even anxiety and depressive moods following SCI. The enhancement of this therapeutic effect suggests that improving the normal excitability of motoneurons is fundamental to alleviating these symptoms.

The 5HT2 receptor is a seven-transmembrane G protein-coupled receptor (GPCR) that interacts with G proteins to stimulate phospholipase C (PLC), leading to increased Ca²⁺ levels.^31,32 5HT2CR is widely distributed in the CNS and has been the focus of research for the treatment of various diseases. However, intriguingly, the effects of 5HT2CR agonists, antagonists, or inverse agonists have shown inconsistency across multiple studies. For example, SERT, a classic antidepressant, is well-known for its role in mood regulation. Both 5HT2CR agonists and antagonists have been proposed for the treatment of anxiety and depression.^33,34,35,36 5HT2CR inverse agonists and agonists are both considered promising for treating spasticity following SCI. The constitutive activity of 5HT2CR and mRNA editing may account for this potential.

mRNA editing and constitutive activity of 5HT2CR are two important features that influence its function. mRNA editing of 5HT2CR supports the diversity of 5HT2CR and determines serotonin signaling and drug therapeutic effects.^37,38 Constitutive activity of 5HT2CR is also related to the extent of RNA editing. RNA editing of 5HT2CR inhibits its constitutive activity.^39,40 The constitutive activity of 5HT2CR allows it to be activated even without ligand binding, meaning its activity is not influenced by agonists. AAV viruses carrying 5HT2C mRNA were injected into SCI mice. We speculate that this treatment may have had an inhibitory effect on the constitutive activity of 5HT2CR.

The effect of 5HT2CR on the relief of spasticity and motor function

5HT2CR can form a complex with GluNR2A, activating Src protein phosphorylation and regulating the depolarization of motor neurons.⁴¹ Additionally, postsynaptic density protein-95 (PSD-95) has been shown to be involved in the signaling pathway of 5HT2CR. Excitotoxicity mediated by NMDA-PSD95-nNOS severely impacts neuronal function after injury.⁴² Studies have shown that fluoxetine, by reducing RNA editing of 5HT2CR, can specifically block glutamate receptors and thereby exert neuroprotective effects.^43,44 We confirmed that 5HT2CR is overexpressed in motor neurons and inhibitory interneurons, significantly reducing the expression levels of NR1 and NR2A in the lumbar enlargement of the spinal cord. This result is consistent with the improvement in motor function observed in our behavioral experiments. Therefore, we hypothesize that the overexpression of 5HT2CR in motor neurons or GABAergic interneurons may be related to the improvement in motor function and the alleviation of spasticity, potentially through the downregulation of NMDARs in the anterior horn of the lumbar spinal cord.

NMDAR plays a key role in regulating neuronal excitability and synaptic plasticity. After SCI, NMDAR may undergo excessive activation, contributing to pathological processes such as excitotoxicity, abnormal spinal cord circuit remodeling, cellular oxidative stress, and inflammatory responses.⁴⁵ The reduction of NMDAR1 and NMDAR2A in the ChAT and VGAT groups likely diminishes excessive excitatory signaling, thereby alleviating spasticity. This is supported by the improved motor function observed in these groups. Although overexpression of 5HT2CR in glutamatergic inhibitory interneurons resulted in some improvement in spasticity in mice, the gait data from the Vglut-2 group did not show significant improvement in walking function. This may be related to the excitatory effects of glutamatergic signaling in spinal reflex regulation. After SCI, the excessive activation of glutamatergic signaling exacerbates the excitability of neural circuits. Motor function, especially coordinated gait, requires a fine balance between excitatory and inhibitory signaling.

It is widely recognized that motor neurons, as the final common pathway for voluntary movement, play a direct role in regulating muscle activity.⁴⁶

The significant benefits observed in the ChAT group suggest that enhancing 5HT2CR signaling in motor neurons may optimize motor output and directly improve gait dynamics. In contrast, the sluggish response in the Vglut-2 group indicates that simply modulating glutamatergic signaling may be insufficient to effectively promote gait recovery. This may stem from the complexity of the glutamatergic network, where excessive activation could disrupt rather than restore motor function, or from compensatory mechanisms that counteract the therapeutic effects of 5HT2CR overexpression.

In summary, the ChAT and VGAT groups highlight the important role of 5HT2CR modulation in motor neurons and GABAergic interneurons in influencing motor output and synaptic plasticity. While glutamatergic signaling is crucial for modulating spasticity, its role in gait recovery is more complex compared to the GABAergic and motor neuron-specific mechanisms. This may involve combinatorial approaches that balance excitation and inhibition while optimizing motor neuron performance. Further studies could provide a more holistic strategy for restoring gait function in spinal cord injury patients.

The expression of 5HT2CR correlates with the regulation of mood and motor function

The results of the OFT study reveal the differential effects of 5HT2CR overexpression in different neuronal subtypes on anxiety and depression-like behaviors in SCI mice. It was found that the ChAT group mice exhibited stronger exploratory behavior, suggesting alleviation of anxiety and depression-like symptoms. However, in the other two groups of mice, especially the Vglut2 group, the improvement in behavior was not as pronounced. Analysis of gait parameters indicated that 5HT2CR overexpression in glutamatergic interneurons did not have an ideal regulatory effect on the mice’s motor function. Perhaps the insufficient improvement in motor function led to the lower central area traversal speed observed in the Vglut group during the OFT. This may suggest that the improvement in the mice’s anxiety and depressive-like behaviors is influenced by the recovery of motor function. In other words, better motor function helps alleviate the anxiety and depression-like behaviors in mice.

Interneurons regulate the activity of neural circuits by releasing inhibitory or excitatory neurotransmitters.⁴⁷ Glutamatergic interneurons are primarily responsible for excitatory signaling within the spinal cord and in ascending spinal regions. Overexpression of 5HT2CR in these neurons may lead to excessive activation of excitatory neural pathways. The imbalance between excitatory and inhibitory signaling in spinal cord neural circuits is closely linked to motor dysfunction, spasticity, and neuropathic pain following SCI.⁴⁸ Although they play an important role in the regulation of neural signaling, their impact on mood regulation is more complex and indirect.

As an SSRI, the activation of 5HT2CR carries the possibility that it not only indirectly modulates mood through its effects on spasticity or pain, but also interacts with neural circuits involved in mood regulation. The serotonergic system is closely linked to these mood-regulating pathways.⁴⁶ The activity of motor neurons may also influence the release of neurotransmitters, such as the serotonergic system itself, or other neurotransmitter systems like dopamine and norepinephrine, thereby having a direct impact on mood.

In summary, the overexpression of 5HT2CR in motor neurons may improve anxiety and depression-like behaviors in SCI mice through direct and/or indirect mechanisms. However, the direct impact of 5HT2CR overexpression on these behaviors requires further investigation. Additionally, we are aware that AAV vectors may be taken up by brain neurons. A study by Kawabata et al. suggests that AAV2 capsid variants predominantly target cerebral vascular endothelial cells, with a lower transduction efficiency in brain neurons.⁴⁹ We believe that the significant impact on 5HT2CR expression in brain neurons and its potential influence on behavioral outcomes is unlikely. However, we also recognize the importance of acknowledging this limitation in the study. In future research, optogenetic manipulation or fiber-optic recording techniques may be used to more directly observe the relationship between motor neuron activity and mood changes.

5HT2CR overexpression in glutamatergic excitatory neurons alleviats neuropathic pain

Mice in the Vglut-2 group were observed to experience milder neuropathic pain, an effect that was not significant in the other two groups. This may be due to the excessive activation of 5HT2CR in glutamatergic excitatory interneurons, which inhibits the presynaptic release of glutamate, reducing the overactivation of pain-related neural circuits. It is reported that inhibit the expression and function of different NMDAR subtypes can alleviate neuropathic pain after SCI.^45,50 Increasing the expression of KCC2 after SCI has also been considered a potential strategy for alleviating neuropathic pain.⁵¹ However, these studies did not focus on the diverse roles of different neuronal subtypes in the onset and treatment of pain.

Our study may provide support for the idea that targeting and modulating 5HT2CR in glutamatergic excitatory neurons is a promising therapeutic strategy for treating neuropathic pain after SCI. Future research should further explore the underlying mechanisms to identify treatment approaches that enhance analgesic effects.

In summary, overexpression of 5HT2CR in glutamatergic excitatory interneurons (Vglut-2 group) can alleviate neuropathic pain following SCI, highlighting the potential of this neuronal subtype for pain treatment.

The therapeutic effect of 5HT2CR overexpression is primarily mediated by modulating GABAAR and KCC2, rather than the BDNF-TrkB pathway

Disruption of chloride homeostasis in motor neurons is thought to be closely linked to the development of spasticity and neuropathic pain after SCI. Research indicates that boosting KCC2 activity can help alleviate hyperreflexia and spasticity associated with chronic SCI.⁵² The BDNF-TrkB pathway plays a crucial role in the regulation of KCC2 and the recovery of neuronal function.^53,54 Moreover, the activation of the BDNF-TrkB pathway and the upregulation of KCC2 expression play key roles in regulating pain hypersensitivity after SCI, contributing to the alleviation of neuropathic pain.⁵⁵ However, our study revealed that overexpression of 5HT2CR in various types of neurons does not significantly affect the BDNF-TrkB pathway in the spinal cord, and KCC2 expression is suppressed. Nevertheless, this observation is crucial for understanding the scope of the therapeutic effects regulated by 5HT2CR. It may suggest that the overexpression of 5HT2CR in specific neurons can improve motor function and reduce spasticity, but this is not mediated by the BDNF-TrkB pathway, which is typically associated with neural plasticity and neuronal recovery. Surprisingly, results from the ChAT and VGAT groups suggest that overexpression of 5HT2CR in motor neurons or GABAergic interneurons might enhance KCC2 function specifically in motor neurons.

As previously mentioned, 5HT2CR, as a GPCR, activates PLC to regulate intracellular calcium ion levels. Additionally, changes in KCC2 and GABAAR are crucial for the regulation of intracellular chloride ion concentration. Proper Cl⁻levels also play an important role in maintaining the function of KCC2 and GABAAR. The expression and function of postsynaptic GABA receptors also regulate the activity of voltage-gated calcium channels (Cav), thereby influencing the excitability of motor neurons.⁵⁶ After SCI, the hyperpolarizing effect of GABAAR on neurons is reduced, leading to abnormal ion plasticity and promoting the development of spasticity and chronic pain.³⁰ This has been shown to be associated with the downregulation of KCC2.⁵⁷ The primary role of KCC2 is to maintain a low level of Cl⁻ concentration within neurons. This function is influenced by GABAAR and KCC2 can also regulate GABAergic signaling and its receptors. In mature neurons, GABAAR triggers synaptic inhibition, whereas in immature neurons, it induces synaptic excitation. Enhancing the function of KCC2 can facilitate the efflux of intracellular Cl⁻, leading to a hyperpolarizing shift in GABAergic activity, which aids in limiting the occurrence of neuronal damage.^58,59,60,61 Serotonergic fibers have a significant impact on the GABAergic system and KCC2.²¹ We found that by increasing the expression of 5HT2CR after SCI, both VGAT and ChAT groups of mice exhibited upregulation of KCC2 at 56dpi, accompanied by downregulation of GABAAR. This suggests that overexpression of 5HT2CR in motor neurons or GABAergic interneurons can enhance KCC2 function in the lumbar enlargement of the spinal cord while suppressing the expression and function of GABAAR, thereby reversing the excessive excitability of motor neurons following injury. The observed upregulation of KCC2 and downregulation of GABAAR in the ChAT and VGAT groups may enhance inhibitory signaling and reduce neuronal hyperexcitability. These molecular changes are consistent with the reduction in spasticity and improvement in motor function observed in these groups. These findings refine our understanding of the molecular targets of 5HT2CR in the context of SCI recovery.

Given the limited efficacy and potential adverse effects of current pharmacological treatments for spasticity and neuropathic pain after SCI, further investigation into targeting 5HT2CR is considered to have important clinical implications. Therapeutic agents targeting 5HT2CR, including agonists, antagonists, and inverse agonists, have been shown to demonstrate some promise in preclinical models. However, these compounds are often characterized by low specificity and are associated with unwanted side effects when administered over extended periods. In contrast, gene therapy has been proposed as a potential strategy for providing more durable and precise treatments. Unlike small-molecule drugs, gene therapy is able to selectively target specific neuronal populations, thereby minimizing off-target effects on other systems. This targeted approach is believed to offer substantial advantages in optimizing therapeutic outcomes and reducing the risks of systemic complications, underscoring the potential of gene therapy as a more effective strategy for SCI management.

To sum, in this study, we demonstrate that 5HT2CR has significant therapeutic potential for alleviating spasticity and related complications following SCI. Specifically, overexpression of 5HT2CR in motor neurons not only significantly improves motor function but also effectively reduces anxiety and depression symptoms. However, this approach shows limited effectiveness in improving mechanical and thermal pain. In contrast, overexpression of 5HT2CR in interneurons, particularly in Gluergic interneurons, provides better relief from neuropathic pain. Our research offers new theoretical support for 5HT2CR as a drug target for chronic complications following SCI and provides a new perspective to address past controversies regarding the function of 5HT2CR. Given the limited efficacy and potential adverse effects of existing medications for treating spasticity and neuropathic pain after SCI, further research on 5HT2CR holds significant clinical relevance.

Limitations of the study

Therapeutic effects of 5HT2CR modulation after SCI vary due to differences in constitutive activity, mRNA editing, and neuronal subtype targeting. While AAV-mediated overexpression offers cell-type specificity, the Cre-LoxP strategy enabled selective expression in motor, GABAergic, and glutamatergic neurons, yet variability in viral transduction efficiency and the possibility of off-target or supraspinal uptake cannot be entirely excluded. Improvements in anxiety- and depression-like behaviors may partly reflect secondary benefits from motor recovery, complicating attribution of direct serotonergic modulation. Finally, future studies integrating electrophysiological or optogenetic approaches will be beneficial for establishing causal links between targeted 5HT2R modulation, circuit dynamics, and functional recovery.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr. Liang-jie Du (duliangjie-ccmu@foxmail.com).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request.
• •This study does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

We thank https://Biorender.com for its technical support. We thank the Beijing Municipal Natural Science Foundation (Grant No. 7212140) for funding.

Author contributions

Conceptualization, H.G. and Z.T.; methodology, H.G. and Z.T.; formal analysis, H.G.,Y.-Z.S., and Z.-M.Y.; writing – original draft, H.G.; writing – reviewing and editing, H.G. and Z.T.; visualization, H.G.; supervision, X.-X.D. and L.-J.D.; project administration, L.-J.D.; funding acquisition, L.-J.D.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit monoclonal anti-BDNF	Abcam	Cat#ab108319, RRID:AB_10862052
Rabbit monoclonal anti-TrkB	Abcam	Cat#ab187041, RRID:AB_2892613
Rabbit monoclonal anti-KCC2	Abcam	Cat#ab259969, RRID:N/A
Rabbit monoclonal anti-ChAT	Abcam	Cat#ab181023, RRID:AB_2687983
Mouse monoclonal anti- GABA A Receptor alpha 2/GABRA2	Abcam	Cat#ab193311, RRID:AB_2890213
Rabbit monoclonal anti-GAD65+GAD67	Abcam	Cat#Ab183999, RRID:AB_2890213
Rabbit recombinant monoclonal anti-VGLUT2	Abcam	Cat#ab216463, RRID:AB_2893024
Rabbit polyclonal anti-5HT2CR	Sigma-Aldrich	Cat#SAB4501477, RRID:AB_10744628
Goat Anti-Mouse IgG Secondary HRP	Bioss	Cat#bs-0296G-HRP, RRID:AB_10893927
Goat Mouse IgG Secondary HRP	Bioss	Cat#bs-0296G-HRP, RRID:AB_10893927
Goat Anti-Rabbit IgG (H + L)	ServiceBio	Cat#GB23303, RRID:AB_2811189
Cy3 conjugated Goat Anti-mouse IgG(H + L)	ServiceBio	Cat#GB21301, RRID:AB_2923552
Cy5-labeled goat anti-mouse IgG	ServiceBio	Cat#GB27301, RRID:AB_2910251
Alexa Fluorescent 488-conjugated Goat Anti-Rabbit IgG(H + L)	ServiceBio	Cat#GB25303, RRID:AB_2910224
Bacterial and virus strains
pHBAAV-CAG-DIO-MCS-T2A-ZsGreen	Hanbio Biotechnology	https://www.hanbio.net/productclass_8/productdetail_11.shtml
Experimental models: Organisms/strains
C57BL/6 N mice	Beijing Vitong Lihua Laboratory	https://www.vitalriver.com/#/animalModel/conventionalAnimal?id=46&idd=9&namecode=productserve&title
B6;129S6-Chattm2(cre)Lowl/J mice	Jackson Laboratory	Strain #:006410 RRID:IMSR_JAX:006410
B6J.129S6(FVB)-Slc32a1tm2(cre)Lowl/MwarJ	Jackson Laboratory	Strain #:028862 RRID:IMSR_JAX:028862
B6J.129S6(FVB)-Slc17a6tm2(cre)Lowl/MwarJ mice	Jackson Laboratory	Strain #:028862 RRID:IMSR_JAX:028862
Software and algorithms
GraphPad Prism 7	N/A	https://www.graphpad.com/scientificsoftware/prism/
SPSS 26.0	N/A	https://www.ibm.com/products/spss-statistics
LabChart 8.0	AD Instruments	https://www.adinstruments.com/support/LabChart

Experimental model and study participant details

Mouse strains and genotyping

All animal experiments were conducted according to protocols approved by the animal care and use committee at Capital Medical University of Science and Technology(Ethics batch No. AEEI-2021-145). Adult female mice weighing 20-25g were used in this study. After arrival, all mice were acclimated to the animal facility at least 7 days before experiments were initiated. Mice were housed in group as 5 mice per cage and kept in a 12-h light/12-h dark cycle at 20°C–25°C room temperature with water and food ad libitum. Only healthy and immunocompetent mice were used. Mice in this study were not involved in any previous procedures.

B6;129S6-Chat^tm2(cre)Lowl/J mice, which express Cre recombinase in cholinergic neurons, without tdisrupting endogenous Chat expression, are referred to as ChAT-IRES-Cre. B6J.129S6(FVB)-Slc32a1^tm2(cre)Lowl/MwarJ mice, which express Cre recombinase in inhibitory GABAergic neuron cell bodies, without disrupting endogenous vesicular inhibitory amino acid transporter expression, are referred to as c. B6J.129S6(FVB)-Slc17a6^tm2(cre)Lowl/MwarJ mice, which express Cre recombinase in glutamatergic neuron cell bodies, without disrupting endogenous vesicular glutamate transporter-2 expression, are referred to as Vglut2-IRES-Cre. All transgenic animals were backcrossed to C57BL/6N for several generations. Heterozygous Cre⁺ mice were used in experiments. Standard genotyping primers are available on the Jackson Lab website. Adult male C57/BL6N mice of SPF grade were obtained from Beijing Vitong Lihua Laboratory Animal Technology Co., Ltd. (Beijing, China).

All animals of different strain were randomized using the random number table. The assessments persons were blinded to all the groups and all outcome analysis was carried out by independent investigators blinded to the treatment condition.

Method details

Animal surgery and post-surgery care

The mice were anesthetized with isopentane (3%–4% induction, 1%–2% maintenance in oxygen). Until the mice entered deep anesthesia, the limbs of mice were fixed in a prone position on the operating table. Fur was shaved, after skin preparation and disinfection, a longitudinal skin incision was made to separate fascia and muscle before exposing the spinal processes. The spinous process and lamina of T10 verteble was removed, spinal cord was fully exposed. The thoracic 9 and 11 vertebra were fixed with a spinal cord impactor (Precision Systems and Instrumentation IH spinal cord impactor, United States) and the exposed spinal cord was impacted with a force of 70 ± 5 kdyn. C57BL/6 N mice were randomly divided into Sham group and SCI group.SCI group accepted the same operation. While the sham group, only laminectomy was performed without hit. After cleaning the operation area, muscle, fascia and skin were sutured layer by layer. At the same time, mice were given 2 × 10⁴U penicillin intraperitoneally to prevent postoperative infection. Mice were placed on a warm blanket throughout operation until wakeup. Five days after surgery, the mice were given penicillin (2 × 10⁴U/day) to prevent infection. The mice were assisted with urination at least once per 12h until the autonomous urination function was restored.

Vector Administration:Tail vein injections

Conditional expression of target genes in Cre-containing neurons was achieved with recombinant adeno-associated viruses (AAVs) encoding a double-LoxP inverted open reading frame (DIO) of target genes. AAV2 is more effective than AAV9 in transducing cell types in the central nervous system.⁶² For chemogenetic experiments with projection targeting, ChAT-IRES-Cre mice(n = 6), Vgat-IRES-cre mice(n = 6),Vglut2-IRES-cre mice(n = 6) were transduced with HBAAV2/BBB-CAG-DIO-m-Htr2c-3xflag-ZsGreen (15×10¹²vg/mL, HanBio,Shanghai, China),divided as ChAT group,Vgat group and Vglut2 group respectively. AAV was diluted with 0.9% saline to a concentration of 1.5 × 10ˆ10 vg/mL before used. 2 days after the SCI operation, we injected AAV. A total of 20 μL of AAV virus solution was injected via tail vein,and the needle was retained 30 s before withdrawing. For C57/BL6 N mice, we injected 20 μL of 0.9% saline 2 dpi after the SCI operation as SCI group(n = 6). And the Sham group(n = 6),we did not give them injection.Coordinates were slightly adjusted based on mouse age and size. Viral vectors were allowed to express for at least four weeks before behavioral evaluation. Furthermore, seven mice were excluded from the study due to mortality and spinal cord impact deviation.

Behavioral tests

Behavioral tests were conducted after SCI surgery. At −7 dpi, which means 7 days before SCI surgery, the pre-training assessment will be used to provide the mice with pre-conditioning.We did RDD by electromyography to test spasticity-like behaviors at 35dpi to 56dpi weeks after transplantation. Basso mouse scale (BMS) was measured at 35dpi and 56dpi, DigiGait footprint analysis was measured at 56dpi for gait analysis. Open field test before was performed at 56dpi. Thermal hyperalgesia testand VonFrey test were performed at 55dpi and 56dpi, respectively.

In vivo electromyography recordings (H-reflex)

The mice were anesthetized with isopentane (3%–4% induction, 1% to 2% maintenance in oxygen).

Two needle electrodes were inserted into the inner side of the lower limb ankle for stimulating the tibial nerve. Two needle-like recording electrodes were inserted into the sole of the foot, a ground electrode was attached to the end of the tail. PowerLab software LabChart (version 8.0, AD Instruments, New South Wales, Australia) was used to convert electrical signals and obtain data. By amplifying the signal (×4000) and filtering it within the range of 300-6k Hz, we stimulate the tibial nerve with a voltage frequency of 0.1Hz and a pulse width of 0.2 ms. We measured the maximum H-wave can be achieved with a current of 1 mA. Then, to measure the RDD of the H reflex, with a 2-min interval,20 consecutive stimuli of 0.2, 0.5, 1, 2, and 5 Hz were performed, respectively. The first 5 H-waves of in each group at different stimulus frequencies were discarded, instead of retaining the waves obtained by last 15 stimuli to determine the RDD at different stimulus frequencies.^63,64,65

DigiGait footprint analysis

Evaluation of hind limb motor function in the SCI mice was assessed by digital footprint analysis system (DigiGaitTM, Mouse Specifics, Inc., Massachusetts, United States). All mice were allowed to walk on the motor-driven belt. The running speed was slowly increased to 5 cm/s. And the foot-prints and body movement were recorded with a high-speed digital video camera from the ventral view of the treadmill belt reflected off the mirror. (Basler high speed camera, Ahrensburg, Germany). Three consecutive gait cycles were evaluated. Three images were recorded for each rat and evaluation.

Basso mouse scale (BMS)

Mice were trained and adapted to the open field before the experiment. The BMS were assessed 35dpi and 56dpi after injury for knowing the locomotor function of mice.²⁷ Each examination was conducted by two experienced observers blinded to the group identity. The two researchers performing and recording the test were unaware of the experimental groupings.

Open field test (OFT)

The exploratory activity and anxiety-like behavior of the mice were evaluated using a Cleversys TopScanLite device (Cleversys, VA, USA).⁶⁶ During the evaluation, the mcie were placed in the center of the open field (40 × 40 cm), the average velocity, distance traveled and time spent in the center zone by each animal was recorded for 5 min using a high-speed camera. Before the second test, the field area were wiped clean and disinfected with alcohol to prevent the odor of the previous mice from affecting the results.

VonFrey filament test

The Von-Frey filament test (Aesthesio, RWD, Danmic/USA) was conducted to assess mechanical allodynia in mice at two time points: before and 8 weeks after SCI. Investigators were blinded to the experimental groupings and treatments. Every mouse was placed in a transparent glass box with a raised wire mesh grid and allowed to adjust to the testing environment for 30 min. Then, mechanical thresholds were measured by applying a Von-Frey wire (0.04–2.0 g) to the lateral plantar surface of each hind paw. A positive reaction was recorded when the foot retracted, the mice licked, or jumped. The left and right hind limbs underwent six tests, each separated by a 5 min interval, and the results were then averaged.

Thermal hyperalgesia test

The Thermal hyperalgesia test was tested to measure thermal injury threshold 1 day after the Von-Frey filament test by a hot plate test (Bioseb, Pinellas Park, FL, United States). The equipment was calibrated to emit heat toward the heat source for determining the baseline withdrawal latency of the animals, which was around 10 s. To prevent scalding, the latency limit was established at 30 s. Positioned beneath the mid-plantar region of the hind paw via a heat conduction plate, a 40W infrared heat source was used. The duration between the start of the stimulus and paw withdrawal was noted for each test. Each mice underwent three evaluations, with each evaluation spaced 10 min apart, and the mean value was calculated from the results.

Western blot analysis

Mice spine samples were removed under deep anesthesia with sodium pentobarbital. Then freezing it with liquid nitrogen and stored at −80°C for western blotting. The tissues were used to detect the expression levels of BDNF (ab108319, Abcam, Cambridge, United Kingdom; dilution 1:1000), TrkB(ab187041, Abcam, Cambridge, United Kingdom; dilution 1:5000), GABAAα2 (ab193311, Abcam, Cambridge, United Kingdom; dilution 1:1000) and KCC2(ab25969, Abcam, Cambridge, United Kingdom; dilution 1:1000). Tissues were centrifuged at 12,000 rpm for 10 min at 4°C after splitting. A BCA protein concentrationassay kit (PC0020, Solarbio, Beijing, China) was used to determine protein concentrations. According to the results,5×loading buffer and PBS were used to adjust the protein concentrations. Then we conducted SDS-PAGE electrophoresis (P0015L,Beyotime,China) and transferred separated proteins to a PVDF (IPVH00010, Millipore, USA) membra. Membranes were then probed with primary antibodies overnight at 4°C. The following morning, the membranes were washed three times, incubated with a secondary antibody (bs-0295G-HRP, bs-0296G-HRP, Bioss, Woburn, MA, USA dilution 1:3000), washed, and analyzed by chemiluminescence and gel image analysis.All mice were euthanized in accordance with relevant animal ethics guidelines and regulations, ensuring compliance with animal welfare standards.

Tissue processing and immunofluorescence

After deeply anesthetized with sodium pentobarbital, the mice quickly entered an anesthetic state and passed away peacefully within a few minutes to minimize pain and distress. The mice were transcardially perfused with 0.9% normal saline.Then, 4% paraformaldehyde solution were used for fixation. Spinal cords from L2 to L6(the lumbar enlargement)were dissected out, post-fixed at 4°C overnight then sectioned on a freezing microtome at a thickness of 30 mm. The tissues designated for staining were outlined using a histochemical pen. Subsequently, the sections were treated with 3% hydrogen peroxide at room temperature for 25 min in the dark, followed by three washes with PBS (pH 7.4). Blocking was carried out using a small amount of 3% BSA before proceeding to incubate the sections with primary antibodies after removing the blocking solution. The primary antibodies including: Chat (ab181923, Abcam, Cambridge, United Kingdom; dilution 1:200), BDNF (ab108319, Abcam, Cambridge, United Kingdom; dilution 1:1000),TrkB(ab187041, Abcam, Cambridge, United Kingdom; dilution 1:5000), GABAAα2 (ab193311, Abcam, Cambridge, United Kingdom; dilution 1:5000) and KCC2(ab25969, Abcam, Cambridge, United Kingdom; dilution 1:5000),Vglut2(ab216463, Abcam, Cambridge, United Kingdom; dilution 1:3000), GAD65 + 67(ab25969, Abcam, Cambridge, United Kingdom; dilution 1:5000), NMDA1R(ab109182, Abcam, Cambridge, United Kingdom; dilution 1:5000), NMDA2AR(ab174636, Abcam, Cambridge, United Kingdom; dilution 1:5000), 5HT2CR(SAB45014, Abcam, Cambridge, United Kingdom; dilution 1:1000). After rewarming and incubation with the corresponding. HRP-labeled secondary antibody, the sections were counterstained with DAPI, washed with PBS, and sealed with an autofluorescence quenching agent (G1221, ServiceBio, Wuhan, China). The tissue sections were imaged under a fluorescence microscope (Eclipse C1, Nikon, Tokyo, Japan) and analyzed by strata Quest (Tissue Gnostics, version 7.0.1.176).

Quantification and statistical analysis

The data are presented as mean ± standard error, and statistical analysis was performed using SPSS software (version 26.0, IBM, Armonk, NY, United States). Two independent sample t-tests were employed for pairwise group comparisons, while one-way ANOVA was utilized for comparisons among multiple groups, with Bonferroni’s post hoc test for correction.For data that do not follow a normal distribution, non-parametric tests were employed: Mann-Whitney U test (for two-group comparisons) and Kruskal-Wallis test (for multiple group comparisons). Statistical significance was defined as p < 0.05.

Published: September 2, 2025