Adult female Sprague–Dawley (SD) rats weighing 250–300 g were obtained from the Experimental Animal Center of Ningxia Medical University. All animal experiments were approved by the Animal Research Ethics Committee of Ningxia Medical University. Animals were kept in a specific pathogen-free environment under a 12-h/12-h light/dark cycle with free access to food and water. After 1 week of adaptation to the environment, these rats were randomly assigned to four groups: Sham group, SCI group, Vehicle group, and Mino group (minocycline, microglia inhibitor). Sham group, in which animals received only laminectomy without SCI; SCI group, in which rats underwent laminectomy and spinal cord contusion; Mino group, in which SCI rats received minocycline (45 mg/kg, diluted in sterilized water, Yuanye Bio-Technology, China) intraperitoneally weekly (Zhang G. et al., 2019; Feng et al., 2021); and Vehicle group, in which SCI rats received the corresponding dose of sterilized water (see Figure 1A).

The rat model of SCI was generated, as previously described (Wu et al., 2014; Chang et al., 2021), with some modifications. Briefly, rats were anesthetized with 1% pentobarbital (1 g/100 ml) and hair was removed from the operation area. Once the T9–T10 segments were located, the skin and fascia were cut along the dorsal midline and the skin was pulled open with a hook to reveal the back muscles. The bilateral paraspinal muscles were sequentially separated to expose the T8–T11 vertebrae, following which the lamina was carefully removed and the T9–T10 segments were fully exposed. The head of the percussion device was aligned with the exposed spinal cord, and a 10-g weight was dropped from a height of 25 cm to the tail end of the percussion device, causing spinal cord contusion by force conduction (Supplementary Figure S1). Hind limb twitching and involuntary tail swinging indicated successful modeling. Animals were well cared for after surgery, with special attention paid to the bladder and rectum.

Newborn SD rats were sacrificed by inhaling excess isoflurane. After removing the skull, the pia meninges and blood vessels were stripped in the precooled D-Hanks buffer. Cortical tissues were separated from the brain and rinsed again and cut into small pieces. Then, these tissues were incubated in Trypsin-EDTA digestion solution at 37°C for 15–20 min. Observation was done every 5 min until the tilted surface was transparent, 10% Dulbecco's modified Eagle medium (DMEM) was added to terminate the digestion. After centrifugation (1,000 RPM, 5 min), the supernatant was removed and the resuspended cells were filtered through a 70-μm filter, yielding a single-cell suspension.

Primary microglia-conditioned medium was harvested from each group, and the levels of IL-18 (EK0592), IL-1β (EK0393), TNF-α (EK0526), and IL-6 (EK0412) were evaluated using associated enzyme-linked immunosorbent assay (ELISA) kits (BOSTER, Wuhan, China), according to the manufacturer's instructions. The absorbance was measured at a wavelength of 450 nm.

Cell viability was measured using a Cell Counting kit (TransDetect, Beijing, China), according to the manufacturer's protocol. Briefly, primary microglia were seeded in 96-well plates. After adhesion, different doses of minocycline (10, 25, 50, 75, and 100 μM) were added to the wells for 12 h. After treatment (minocycline or microglia-conditioned medium), complete medium containing 10% cholecystokinin (CCK) was added to the wells for 1 h. The absorbance was assessed at a wavelength of 450 nm.

Primary neurons cultured in microglia-conditioned medium were centrifuged (4°C, 300 × g, 5 min) and washed two times with precooled PBS, and the cell density was adjusted to 1 × 10⁶ cells/ml with the addition of 400 μl of Annexin V binding solution. The cells were then incubated for 15 min with Annexin V–FITC staining solution (5 μl) and then for 5 min with propidium iodide (PI) staining solution (5 μl) at 4°C. Finally, the level of apoptosis was determined by flow cytometry.

The rats were deeply anesthetized. The skull was decapitated and stripped off, then the brain could be integrally removed. The pial and blood vessels on the brain surface were removed on ice and placed into a precooled rat brain mold. According to the rat brain atlas, the excess brain tissue was discarded along the corona and a certain region of the M1 was retained. Total protein was extracted from a fresh M1 tissue using the Total Protein Extraction kit (KGP2100, KeyGEN BioTECH, China), and protein concentrations were determined using the BCA Protein Assay kit (KGPBCA, KeyGEN BioTECH, China). Equal amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to 0.2-μm polyvinylidene difluoride (PVDF) membranes (10600021, Amersham, Germany). Then, 5% bovine serum albumin (BSA) was applied to block non-specific binding at room temperature for 2 h. Next, the membrane was incubated with the following primary antibodies at 4°C overnight: NLRP3 (bs-10021R, Bioss, China), ASC (bs-6741R, Bioss, China), caspase-1 p20 (bs-10442R, Bioss, China), IL-1β (bs-0812R, Bioss, China), IL-18 (bs-4988R, Bioss, China), TNF-α (bs-10802R, Bioss, China), IL-6 (DF-6087, Affinity, USA), Active Caspase3 (bsm-33199M, Bioss, China), and β-actin (66009-1-Ig, Proteintech, USA) (all 1:1000). After washing three times in 0.1% tris-buffered saline containing 5% Tween-20, membranes were incubated in appropriate secondary antibodies at room temperature for 2 h: Goat anti-Mouse (1:4000, 926-32210, LI-COR, USA) and Goat anti-Rabbit (1:4000, 926-32211, LI-COR, USA). Finally, these membranes were visualized in an Odyssey Infrared Imaging System (CLX-0796, Gene Company Limited, USA) and quantified with ImageJ software (National Institutes of Health).

Rats were deeply anesthetized and transcardially perfused with 0.9% saline (250 ml, room temperature) and 4% paraformaldehyde solution (500 ml, 4°C). Then, the brains were prepared into 5-μm-thick paraffin sections. After dewaxing and dehydration, the sections were incubated in Nissl staining solution (G1436, Solarbio, China) for 30 min at 50–60°C. Subsequently, these sections were washed with flowing water, dehydrated with 100% ethanol, and cleared with xylene. Finally, stained sections were imaged under a light microscope.

The protocol used to prepare the tissue sections was the same as Nissl staining. Then, the sections were boiled in sodium citrate antigen repair solution for 20 min, and immersed in 0.5% Triton X-100 for 30 min. Also, cultured microglia and neurons were harvested and fixed. The brain sections or cells were blocked in 5% BSA for 2 h at room temperature. Subsequently, these sections or cells were incubated with the following primary antibodies at 4°C overnight: Iba-1 (1:500 brain sections and 1:1000 cells, ab178846, Abcam, UK), NLRP3 (1:400 brain sections and 1:1000 cells, bs-10021R, Bioss, China), ASC (1:400 brain sections and 1:1000 cells, bs-6741R, Bioss, China), caspase-1 p20 (1:400 brain sections and 1:1000 cells, bs-10442R, Bioss, China), NeuN (1:500, ab177487, Abcam, UK), Active Caspase3 (1:400, bsm-33199M, Bioss, China), and MAP2 (1:200, GTX133109, GeneTex, USA). After washing three times in PBS containing 5% Tween-20, brain sections and cells were incubated in appropriate secondary antibodies at 4°C for 8 h: Goat Anti-Rabbit IgG FITC (1:100, SA00003-2, Proteintech, USA) and Goat Anti-Mouse IgG Alexa Fluor 647 (1:100, ab150115, Abcam, UK), after which cell nuclei were stained with DAPI. Finally, brain sections and cells were imaged under a multichannel fluorescence microscope (Leica, Wetzlar, Germany) and quantified with ImageJ software.

Motor-evoked potentials (MEPs) were measured using electrophysiology (Redondo-Castro et al., 2016; Chen et al., 2020). Animals were anesthetized. A suspension silver sphere electrode as an anode was placed in the hindlimb representation of the motor cortex, just touching the dura without compression. The reference electrode as a cathode was placed under the skin between the ears. The recording electrode was inserted into the tibialis anterior muscle, and the grounding electrode was subcutaneously inserted. A single electrical pulse (10 mA, 0.1 ms, and 1 Hz) was used to stimulate the brain and record the amplitude of MEPs.

Locomotor function was assessed using the Basso, Beattie, and Bresnahan (BBB) rating scores and the inclined plate test on days 1, 3, 7, 14, 21, and 28 after the operation (Basso et al., 1995; Gu et al., 2020). Rats were placed in an open field area or an inclined plane. The average BBB rating scale score and the maximum angle of the plane were assessed and recorded by two evaluators blinded to the experimental conditions.

All data were analyzed in GraphPad Prism 6.01 (GraphPad Software, USA) and are presented as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) with the Bonferroni test was applied for comparisons among multiple groups. A p-values < 0.05 were considered significant.

To assess the microglial activation status, we employed immunofluorescence staining to detect the expression of Iba-1, a microglial marker protein, and NLRP3, ASC, and caspase-1 p20 (caspase-1 subunit), components of the NLRP3 inflammasome (Figures 1B–E). Microglial activation was evident in M1 after SCI; however, this effect was blocked by minocycline treatment (an inhibitor of microglial activation) (Chen et al., 2018; Wang et al., 2018; Su et al., 2019). Meanwhile, the average optical density of NLRP3, ASC, and caspase-1 p20 was high after SCI, whereas the opposite was seen after minocycline treatment (Figures 1F–I). Similarly, western blot (WB) analysis revealed that the expression of NLRP3, ASC, and caspase-1 p20 was high after SCI, but this effect was blocked with minocycline treatment (Figures 2A–D). Given that increased NLRP3 inflammasome activation can lead to the maturation of IL-18 and IL-1β, which further promotes the secretion of TNF-α and IL-6, we next detected the expression of IL-18, IL-1β, TNF-α, and IL-6 in M1. As expected, the expression of all four proteins was increased following SCI, but returned to near-normal levels after minocycline treatment (Figures 2E–I). These findings suggested that microglial activation mediated NLRP3-related neuroinflammation in M1 following SCI.

Damaged neurons often undergo apoptosis, and their ability to recover is limited. Determination is made whether an increase in inflammation in M1 caused an injury to neurons. We performed Nissl staining and evaluated the expression of cleaved (activated) caspase-3, a marker of apoptosis. The results showed that the number of Nissl-positive neurons was decreased, while that of NeuN/cleaved caspase-3 double-positive neurons was increased after SCI (Figures 3A–D). Similarly, cleaved caspase-3 protein levels were also increased after SCI (Figures 3E,F). However, both these effects were mitigated by minocycline treatment.

Next, we investigated the functional integrity of the motor pathway using MEPs. The results showed that the amplitude of MEP was markedly reduced after SCI; however, this reduction was reversed following minocycline treatment (Figures 4A,B). Subsequently, the BBB rating scale and inclined plane test scores were determined as a readout of locomotor function. We found that locomotor function was severely curtailed after SCI, whereas the opposite was seen when minocycline was administered (Figures 4C–E). In total, our results suggested that suppressing neuroinflammation in M1 could improve the integrity of the motor pathways and enhance locomotor function recovery after SCI.

We successfully cultured primary cortical microglia and neurons with >95% positive cells of landmark protein (Supplementary Figures S2, S3). To determine the effect of minocycline on cell viability of primary microglia, a dose-response experiment was performed on microglia treated with minocycline for 12 h. Similar to previous studies (Lu et al., 2016; Piotrowska et al., 2017), cell viability was not changed in response to minocycline (10–100 μM) (Figure 5B). And, we chose 50 μM minocycline treatment for 12 h at subsequent experiments. Microglia were activated after LPS+Nig treatment, which were prevented by minocycline pretreatment. Meanwhile, the average optical density of NLRP3, ASC, and caspase-1 p20 was high after LPS+Nig treatment, and the average optical density of these three proteins was low after minocycline pretreatment (Figures 5C–J). The secretion of IL-18, IL-1β, TNF-α, and IL-6 was high after LPS+Nig treatment, which was also restrained after minocycline pretreatment (Figures 5K–N). Thus, we acquired microglia-conditioned medium. We further found that primary cortical neurons cultured in microglia-conditioned medium following LPS+Nig treatment experienced injury. In this group, cell viability and the average optical density of MAP2 were reduced, whereas the average optical density of cleaved caspase-3 and the levels of apoptosis were increased. Furthermore, negative effects associated with the activated microglia-conditioned medium were reversed with minocycline pretreatment (Figure 6). In total, the vitro results demonstrated that microglial activation mediated NLRP3-related neuroinflammation, resulting in damage to cortical neurons.