In pathophysiological terms, SCI is categorized into primary and secondary injury. Temporally, the injury progression can be delineated into the acute phase (within the first 48 h), the subacute phase (spanning from 48 h to 14 days), and the chronic phase (persisting beyond 14 days) (Ahuja et al., 2017; Ribeiro et al., 2023).

3.1 Acute phase (<48 h)3.1.1 Time: 5 min3.1.1.1 Pathological hemodynamic changes

In vivo imaging experiments revealed that pathological blood acceleration occurred around 5 min after SCI, followed by blood stasis or clot removal (de Ruiz Almodovar et al., 2025). Blood components began to leak during the same time period. Since endothelial cells on BSCB are directly exposed to blood flow, pathological hemodynamic changes are one of the factors that lead to increased BSCB permeability and subsequent BSCB disruption (Zhou et al., 2023). The mechanism of pathological hemodynamic changes on BSCB mainly involves two aspects:

Initially, pathological alterations in hemodynamics can impose aberrant mechanical stresses on the vascular wall, leading to its structural compromise (Chiu and Chien, 2011). Subsequently, we computed key hemodynamic parameters affecting the endothelium by applying Poiseuille’s law (Bougouin et al., 2024). The pressure gradient serves as the driving force for maintaining blood flow. Concurrently, it generates an outward force on the vascular lumen, termed transmural pressure, which modulates the passive exchange of water-soluble substances across BSCB (Zhou et al., 2023). Elevated intravascular pressure reflects a heightened outward driving force acting upon compromised vessels, which promotes vascular extravasation, impedes the clearance of tissue fluid, and exacerbates tissue edema (Jacob et al., 2016; Leonard et al., 2015). Under pathological hemodynamic conditions following SCI, a marked elevation in the pressure gradient within the dorsal ascending vein (dAV) was observed. This finding suggests a significant increase in the driving force for permeation across BSCB during this phase (Zhou et al., 2023).

Secondly, the vascular wall is subjected to a physical force generated by blood flow, known as shear stress. Within the range of 10–20 dyn/cm2, shear stress is crucial for preserving endothelial barrier integrity. Existing research indicates that deviations from physiological shear stress—whether excessively high or low—can adversely affect both the morphology and function of endothelial cells. Such alterations result in compromised intercellular junctions and consequent dysregulation of barrier permeability (Garcia-Polite et al., 2017). Previous studies have calculated the shear stress and shear rate in the vessels based on Newton’s law of internal friction, using a simplified method (Tarbell, 2010). The shear stress in the dAV increased significantly after SCI, and was not affected by TTM treatment. There is evidence that in vitro, shear stress above 40 dyn/cm2 can significantly alter junctional structures and reduce the expression levels of TJ proteins (Garcia-Polite et al., 2017). During the rapid and widespread disruption of BSCB after SCI, more than a quarter of the small veins were under shear stress above 40 dyn/cm2, indicating that this adverse frictional force has a wide range of impact.

In summary, pathological hemodynamics after SCI lead to BSCB disruption, increased permeability, inflammatory cell extravasation from vessels, and aggravate the severity of secondary injuries such as inflammation. This is achieved by applying increased pathological transmural pressure and shear stress to the endothelium (Zhou et al., 2023) (Figure 2A).

(A) Pathological hemodynamics after SCI cause BSCB disruption by applying increased pathological transmural pressure and shear stress to the endothelium; (B) Mast cells can secrete chymase, which can convert Big ET-1 into a 31-amino acid intermediate peptide ET-1(1–31). ET-1(1–31) is then hydrolyzed by NEP to produce ET-1. In addition, ECE can directly convert Big ET-1 into ET-1. ET-1 mediates endothelial cell remodeling and causes BSCB disruption through ETA or ETB2 receptor subtypes; (C) Endoplasmic reticulum stress after SCI activates CHOP and Caspase-12, which further leads to endothelial cell apoptosis and ultimately results in the degradation of TJ and AJ proteins. NBP can protect BSCB by inhibiting endoplasmic reticulum stress after SCI.

3.1.2 Time: 30 min3.1.2.1 Increase of endothelin

The concentration of endothelin in spinal cord tissue following moderate injury was measured using an enzyme-linked immunosorbent assay. According to their findings, endothelin levels in the spinal cord exhibited a 50% elevation, which was observed no earlier than 30 min post-injury (McKenzie et al., 1995). Endothelin-1 (ET-1), initially isolated from the culture supernatant of aortic endothelial cells by Yanagisawa et al., is a peptide consisting of 21 amino acids. Following the discovery of ET-1, two further isoforms—endothelin-2 (ET-2) and endothelin-3 (ET-3)—have been characterized. All three isoforms are products of distinct genes, share a common structure of 21 amino acid residues stabilized by two disulfide bonds, and possess six conserved amino acid residues at their carboxyl-terminal ends (Liu et al., 2023).

The endothelin-converting enzyme (ECE) family includes three membrane-bound subtypes: ECE-1, ECE-2, and ECE-3. Among these, the initial two subtypes are capable of being further subdivided into distinct functional variants (Xu et al., 1994). These isoforms exhibit variations at their N-terminal regions and display distinct intracellular localization patterns. The principal role of these enzymes lies in the hydrolysis of the precursor molecule Big ET-1. ECE functions as the key systemic enzyme in the production of ET-1, directly facilitating the maturation of the 38-amino acid precursor into the active peptide (D'Orléans-Juste et al., 2018). The production of endothelin-1 (ET-1) is not exclusively dependent on the ECE-mediated pathway. In embryonic mice with genetic knockout of both ECE-1 and ECE-2, ET-1 levels decreased by only 33% (Yanagisawa et al., 2000). This indicates that there are other pathways to produce ET-1, rather than relying on the typical endothelin converting enzyme. One alternative pathway involves chymase derived from mast cells. This serine protease catalyzes the transformation of Big ET-1 to ET-1. In particular, chymase achieves this by cleaving the Tyr31-Gly32 bond within the 38-amino acid precursor Big ET-1, producing an intermediate 31-amino acid peptide known as ET-1(1–31). Mature ET-1 is then formed both in vitro and in vivo through the hydrolysis of ET-1(1–31) by neutral endopeptidase, which specifically severs the Trp21-Val22 bond (Fecteau et al., 2005).

The pathophysiological effects mediated by endothelin-1 (ET-1) are partially attributable to its receptor subtypes. At least three such subtypes—ETA, ETB1, and ETB2—are known to influence vascular reactivity within the central nervous system. Among them, the ETA receptor subtype shows a particularly strong association with vascular spasm. The ETB receptor category is subdivided into ETB1 and ETB2. The ETB1 subtype, localized on vascular endothelial cells, primarily mediates vasodilation. In contrast, the ETB2 subtype is expressed on vascular smooth muscle cells and predominantly induces vasoconstriction. Consequently, ET-1-induced damage to spinal cord vasculature, including barrier disruption and vasoconstriction, may be predominantly mediated through the ETA or ETB2 receptor subtypes (Dmour et al., 2023). ETA and ETB receptors are widely distributed within the spinal cord. Their associated signal transduction pathways critically influence inflammation and oxidative stress, both of which are recognized as significant factors impacting neural recovery following BSCB disruption induced by SCI (Ranjan and Gulati, 2022). Within the vascular lumen, the signaling actions of ET-1 are mediated by two specific G protein-coupled receptors: ETA and ETB2 (the latter is not shown). Activation of these receptor pathways facilitates endothelial cell remodeling, upregulates adhesion molecule expression, and impairs the integrity of the blood-spinal cord barrier (D'Orléans-Juste et al., 2018) (Figure 2B).

Moreover, injury to the endothelium upregulates ET-1 expression within endothelial cells and activates monocytes to secrete chemokines, including CXCL8. These chemokines facilitate the margination and transendothelial migration of inflammatory cells across the blood–brain barrier. Additionally, macrophages are capable of converting big-ET-1 into ET-1 and simultaneously producing various cytokines and reactive substances. Among these are tumor necrosis factor alpha, interleukin-1, chemokines like CCL5 and CCL2, and reactive oxygen species, all of which are biologically active within the central nervous system (D'Orléans-Juste et al., 2018; McCarron et al., 1993; Schinzari et al., 2024).

3.1.2.2 Endoplasmic reticulum stress

As a principal subcellular organelle of eukaryotic cells, the endoplasmic reticulum (ER) participates in the synthesis and proper folding of secretory and membrane-associated proteins (Kang et al., 2022). This process is critical for maintaining normal cellular function and viability. ER stress denotes a state of disrupted ER homeostasis triggered by diverse exogenous or endogenous stimuli. This disruption results in the accumulation of misfolded or unfolded proteins within the ER lumen, surpassing its intrinsic processing capacity (Zhou et al., 2021). ER stress has been demonstrated to contribute to both the initiation and progression of SCI. This pathological process subsequently results in disruption of BSCB and neuronal death (Hu et al., 2024; Kang et al., 2022; Zheng et al., 2025).

Research indicates that the upregulation of endoplasmic reticulum stress-related gene expression commences within 30 min post-injury and persists for up to 24 h (Li et al., 2021; Piri et al., 2026). After SCI, apoptosis in EC is initiated by the activation of the ER stress-responsive transcription factor C/EBP homologous protein (CHOP) and caspase-12 (Hu et al., 2025). This programmed cell death subsequently promotes the breakdown of proteins constituting TJ and adherens junctions (AJs). Consequently, the integrity of BSCB is compromised, leading to increased permeability. This allows for the infiltration of blood cells, plasma components, and inflammatory factors into the spinal cord parenchyma, thereby contributing to a range of secondary injuries, including inflammatory pathology. N-butylphthalide (NBP) treatment protects BSCB function by inhibiting ER stress after SCI (Zheng et al., 2017) (Figure 2C).

3.1.3 Time: 6 h3.1.3.1 Fluctuation of the transcription factor BMAL1

The transcription factor BMAL1/ARNTL is an indispensable element within the core clock circuitry, governing the circadian oscillations of gene expression (Slomnicki et al., 2020). ARNTL/BMAL1, a basic helix–loop–helix transcription factor (TF), functions in conjunction with its binding partners CLOCK or NPAS2 to modulate the circadian rhythmicity of gene expression (Wang et al., 2024). The core circadian oscillator pathway, comprising transcription factors such as Nr1d1/2, Dbp1, Cry1/2, and Per1/2, is regulated by BMAL1: CLOCK/NPAS2 heterodimers. Additionally, tissue-specific target genes of these heterodimers contribute to the circadian regulation of metabolic processes, immune/inflammatory responses, and antioxidant defense mechanisms (Early et al., 2018; Koike et al., 2012; Musiek et al., 2013; Oishi et al., 2017; Rey et al., 2011). While BMAL1 is essential for normal circadian function, its expression is dynamically altered following spinal cord injury (SCI). Studies have confirmed the expression of BMAL1 in spinal cord ECs and shown that its expression is significantly upregulated at 6 and 24 h post-SCI (Slomnicki et al., 2020). Interestingly, experimental reduction of this upregulated BMAL1 promotes functional recovery after SCI. The primary mechanism for this benefit appears to be the alleviation of neuroinflammation through the reduction of BSCB damage and hemorrhage.

BMAL1 reduction alleviates BSCB damage and neuroinflammation through the following mechanisms: Firstly, BMAL1 reduction can decrease Nos3 levels, and Nos3 expression/activity reduction is also beneficial for BSCB function, because NO produced by Nos3 may cause pathological damage to BBB/BSCB (Beauchesne et al., 2009). Furthermore, given the involvement of neovascularization in the delayed deterioration of BSCB function post-SCI, the downregulation of pro-angiogenic genes within the BBB/BSCB module resulting from diminished BMAL1 may represent an additional mechanism for mitigating this delayed functional impairment (Slomnicki et al., 2020). In summary, although BMAL1 is acutely upregulated after SCI, interventions that reduce its levels can decrease BSCB disruption and neuroinflammation. Therefore, BMAL1 represents a potential therapeutic target for mitigating BSCB impairment and attenuating the inflammatory response following SCI (Figure 3A).

(A) BMAL1 degradation reduces NOS3 levels, limiting NO-mediated BSCB damage, and inhibits pro-angiogenic genes, protecting BSCB; (B) NF-κB and Jmjd3 interact to demethylate H3K27me3 at Mmp promoters, initiating Mmp-3 and Mmp-9 expression. These degrade TJ proteins, disrupting BSCB; (C) Heme leakage through BSCB induces HO-1, limiting vascular adhesion molecules, neutrophil infiltration, and early inflammatory damage. HO-1 directly or via ATF4 inhibition suppresses NLRP1 expression, reducing NLRP1 inflammasome formation and BSCB injury.

3.1.4 Time: 7 h3.1.4.1 Up-regulation of demethylase Jmjd3

Histone modification represents a significant regulatory mechanism influencing gene expression and associated biological processes, including development, metabolism, pathogenesis, and diverse cellular responses (Bannister and Kouzarides, 2011). Moreover, certain investigations indicate that factors governing histone modifications are critically involved in post-spinal cord injury cellular events, such as glial cell dynamics, neuroprotection, inflammatory regulation, blood-spinal cord barrier disruption, and locomotor function (Lu et al., 2013; Lv et al., 2012). The research indicates that following spinal cord injury, there is an elevation in the levels of Jmjd3, a histone H3K27 demethylase enzyme, within endothelial cells. This increase in Jmjd3 expression occurs within 7 h after SCI (Lee et al., 2012a). Jmjd3 triggers neuroinflammation after SCI through two mechanisms: on the one hand, it activates MMP to damage BSCB; on the other hand, it directly promotes the expression of pro-inflammatory mediators, including iNOS, IL-6, TNF-α, COX-2, and IL-1β, in infiltrating inflammatory cells following BSCB disruption, thereby amplifying the neuroinflammatory cascade (Lee et al., 2016). Here are the details:

First of all, Jmjd3 is required to activate Mmp-3 and Mmp-9 genes by demethylating H3K27me3 at the MMP promoter. MMP3 and MMP9 gene activation requires intact NF-κB binding sites. NF-κB functions not only as a transcriptional activator for Jmjd3 but also as a transcription factor that physically interacts with Jmjd3. NF-κB and Jmjd3 cooperate in activating MMP function (Kang et al., 2011). The synergistic increase in Jmjd3 and NF-κB expression is required to initiate the transcription of Mmp-3 and Mmp-9 genes in vascular endothelial cells following injury (Lee et al., 2012a). In summary, Jmjd3 cooperates with NF-κB to promote the transcriptional activation of Mmp-3 and Mmp-9. This is achieved through demethylation of H3K27me3 at the promoters of these Mmp genes. The elevated expression of Mmp-3 and Mmp-9 markedly enhances the permeability of BSCB and contributes to the degradation of tight junction proteins, consequently exacerbating secondary injuries, such as post-SCI inflammation (Lee et al., 2016).

Secondly, Jmjd3 directly upregulates the expression of pro-inflammatory cytokines and enzymes—such as iNOS, IL-6, TNF-α, COX-2, and IL-1β—in immune cells that infiltrate the spinal cord after BSCB breakdown, thus intensifying the local inflammatory response and secondary tissue damage (Lee et al., 2016; Linnerbauer et al., 2020) (Figure 3B).

Furthermore, research indicates that suppressing Jmjd3 expression enhances functional recovery following SCI, while also diminishing axonal degeneration, myelin loss, and the extent of tissue damage. Accumulated evidence indicates that Jmjd3 represents a promising therapeutic target for preventing BSCB disruption following SCI (Lee et al., 2016). Furthermore, the relevance of our results encompasses multiple central nervous system (CNS) disorders associated with BBB/BSCB dysfunction, such as meningitis, epilepsy, Alzheimer’s disease, amyotrophic lateral sclerosis, and multiple sclerosis (Park et al., 2024; Yang and Rosenberg, 2011).

3.1.5 Time: 18 h3.1.5.1 The formation of heme oxygenase-1

Heme Oxygenase-1 (HO-1) is an enzyme responsible for the degradation of heme, yielding metabolites including carbon monoxide, bilirubin, and free iron. These products exert antioxidant, anti-inflammatory, and neuroprotective actions (Wei et al., 2024). A study showed that the formation of HO-1 after SCI may be induced by the extravasation of heme from the disrupted BSCB. HO-1 induction peaked at 18 h after SCI and decreased at 24 h relative to 18 h. HO-1 limited the damage of inflammation to the BSCB after SCI through two ways. First, HO-1 can alter the expression of vascular adhesion molecules, which regulate the blockade and rolling of neutrophils, so the induction of HO-1 can therefore restrict the migration of inflammatory cells. Thereby attenuating the early infiltration of neutrophils into BSCB and mitigating BSCB damage caused by inflammatory factors (Yamauchi et al., 2004). Secondly, HO-1 has been shown to suppress the assembly of the NLRP1 inflammasome following SCI. As a multi-protein complex, the NLRP1 inflammasome serves to activate both cytochrome c and caspase-1, a process that subsequently induces apoptosis. This apoptotic pathway plays a significant role in modulating neuronal death and inflammatory responses after SCI (Hellenbrand et al., 2021). Experimental findings indicate that the NLRP1 inflammasome is detectable within the cerebrospinal fluid of patients with SCI and may contribute to impairment of BSCB (de Rivero Vaccari et al., 2016). Different types of inflammatory corpuscles not only exist in inflammatory immune cells, but also in neurons. There is increasing evidence to support this notion. It has been demonstrated that NLRP1 inflammasome can induce neuronal damage under high glucose levels (Meng et al., 2014). Activation of Panconnexin channels by high extracellular potassium leads to inflammasome activation in primary neurons and astrocytes (Silverman et al., 2009). Importantly, it should be noted that inflammatory corpuscles have been observed in various types of neurons, including those found in spinal cord injury lesions (de Rivero Vaccari et al., 2012; de Rivero Vaccari et al., 2016; Tan et al., 2015; Walsh et al., 2014). HO-1 can inhibit the formation of NLRP1 inflammasome by down-regulating the expression of NLRP1 in neurons, thus protecting neurons from SCI-induced injury. Furthermore, HO-1 reduces NLRP1 transcription by inhibiting the expression of activating transcription factor 4 (ATF4). This transcription factor, which is upregulated under cellular stress conditions, regulates NLRP1 gene promoter activity and contributes to the increased NLRP1 expression observed after SCI (Lin et al., 2016). Therefore, HO-1 has anti-inflammatory effects in the local area after SCI, reducing the damage of inflammation to the BSCB, thus stabilizing the BSCB (Figure 3C).

3.1.6 Time: 1 day3.1.6.1 Bradykinin activates its receptor

The physiological effects of bradykinin are primarily mediated through its binding to specific receptors on the cell surface. Identified as the two major subtypes, bradykinin receptor B1 (B1R) and bradykinin receptor B2 (B2R) belong to the G protein-coupled receptor (GPCR) family (Astuti and Ysrafil, 2020; Yi et al., 2024). Bradykinin and bradykinin receptors peaked at 1 day after SCI and dropped to the lowest at 7 days after SCI (Yang et al., 2021). Following B1 receptor activation by bradykinin, the expression of leukocyte adhesion molecules, such as ICAM-1 and VCAM-1, is upregulated on endothelial cells. This promotes leukocyte-endothelial interaction, facilitating the migration of leukocytes across BSCB into neural tissue. Within the neural parenchyma, these cells release pro-inflammatory mediators such as TNF-α, IL-1β, and IL-6, thereby exacerbating the infiltration of inflammatory cells and factors into the BSCB region (Medeiros et al., 2004; Rex et al., 2022). Bradykinin can also activate B2 receptor, induce neurons and glial cells to release prostaglandin E2 (PGE2), further enhance BSCB disruption and inflammatory response (Brusco and Oliveira, 2025). PGE2 is a cyclooxygenase (COX)-dependent prostaglandin that is synthesized and released in large amounts after SCI. PGE2 exerts its effects by binding to and activating its cognate receptors, EP1 through EP4. This activation modulates the infiltration of inflammatory cells and the release of inflammatory mediators, ultimately intensifying the influx of both inflammatory cells and factors into BSCB (Starikova et al., 2025). In addition, PGE2 can also further induce endothelial cell contraction and tight junction gap formation, reduce BSCB integrity, increase BSCB permeability, facilitate the migration of inflammatory cells from vessels to spinal cord neural tissue (Lee et al., 2020). Therefore, the mechanism of bradykinin in BSCB disruption after SCI is multifaceted, and bradykinin and its receptors may be potential targets for treating BSCB injury and inflammatory response after SCI (Figure 4A).

(A) Bradykinin activates B1R, upregulating ICAM-1/VCAM-1 to promote leukocyte infiltration and inflammation. Via B2R, it induces neuronal/glial release of PGE2, whose receptors (EP1–4) trigger inflammatory factors, damaging BSCB; (B) Neuron- and macrophage-derived HMGB1 activates astrocytes and microglia via neural pathways, causing inflammation and BSCB damage; (C) Post-SCI DJ-1 upregulation induces SOCS1, inhibiting ROS. ROS activates the NLRP3 inflammasome, promoting maturation/release of IL-18, IL-1β, and caspase-1, creating conditions for MMP-9-mediated TJ degradation; (D) MMP-8 can reduce the expression levels of TJ proteins occludin and ZO-1, thereby causing BSCB disruption.

3.1.6.2 The production of high-mobility group box 1

Tissue analysis from patients with SCI reveals increased expression of High Mobility Group Box 1 (HMGB1). This molecule is implicated in driving neuroinflammation, promoting neuronal death, and facilitating ferroptosis (Kigerl et al., 2018; Ren et al., 2021; Taverna et al., 2022). The HMGB1 protein, consisting of 215 amino acids (aa), is derived from a gene located on chromosome 13q12. The gene structure encompasses 5 exons and 4 introns. Structurally, it contains three domains: two HMG box domains (designated A box and B box) joined by a short flexible linker, and a C-terminal tail characterized by an enrichment of glutamate and aspartate residues spanning approximately 30 aa (Bianchi et al., 1992). The A box domain has HMGB1 antagonistic activity, while the B box has pro-inflammatory activity (Tian et al., 2020). HMGB1 exerts distinct biological functions depending on its subcellular and extracellular localization. This factor can promote the production of pro-inflammatory cytokines, orchestrate cellular activities including proliferation, differentiation, and invasion, and also control autophagic processes (Chen et al., 2022; Yuan et al., 2024).

Following SCI, the upregulation of HMGB1 precedes that of cytokines like TNF-α, IL-1β, and IL-6 (Chen et al., 2011). HMGB1 protein levels in the spinal cord show a marked increase within 12 h to 3 days post-injury, peaking on the first day. This release of HMGB1 is primarily mediated via the HMGB1/RAGE or TLR signaling pathways, originating from neurons and macrophages (Sun et al., 2019). HMGB1 is capable of binding to a variety of receptors, including TIM-3, TLR2, TLR4, TLR9, RAGE, and CXCR4. Following SCI, elevated expression levels of HMGB1, RAGE, and TLR2/4 are observed. This increase facilitates the interaction between HMGB1 and either the RAGE or TLR2/4 receptors (Chen et al., 2011). The HMGB1-RAGE pathway can induce macrophage/microglia polarization to pro-inflammatory phenotype, thereby inducing inflammation (Cavalcanti et al., 2025; Fan et al., 2020). The HMGB1/TLR-4 signaling axis also contributes to the activation of inflammatory responses following SCI (Wang et al., 2021). In addition, HMGB1 can also affect the COX2/PGE2 pathway in SCI, leading to astrocyte inflammation and further aggravating the infiltration and damage of inflammatory cells and inflammatory factors to BSCB (Song et al., 2021). Beyond its role in promoting post-spinal cord injury inflammation, HMGB1 can disseminate to the brain via the bloodstream, cerebrospinal fluid (CSF), and axonal transport. This spread contributes to the impairment of both BSCB and BBB. HMGB1 worsens the situation after SCI and hinders SCI recovery, therefore HMGB1 is a potential therapeutic target (Wu and Li, 2023) (Figure 4B).

3.1.6.3 Increased expression of DJ-1

DJ-1 participates in multiple pathophysiological processes, including oncogenesis, the regulation of mitochondrial function, and the inhibition of protein glycosylation (Clements et al., 2006). Prior research has established that DJ-1 exerts neuroprotective effects in the context of neurodegenerative disorders and cerebral ischemia (Wang et al., 2020). A proposed mechanism underlying this protection involves the modulation of inflammatory processes (Zhao et al., 2022). Within the pathological environment of ischemic stroke, DJ-1 demonstrates anti-inflammatory effects through the reduction of pro-inflammatory cytokine levels, such as TNF-α, IL-1β, and IL-18 (Lin et al., 2024; Nakamura et al., 2021; Peng et al., 2020). Conversely, DJ-1 also confers protection against SCI through modulation of BSCB permeability and inflammatory responses. Studies indicate that DJ-1 expression begins to rise within 3 h post-SCI, peaking at 24 h. After 24 h of injury, the protein level of DJ-1 decreased significantly (Cai et al., 2022).

DJ-1 causes BSCB damage and inflammation through the following mechanisms: According to the study, DJ-1 is upstream of SOCS 1 (cytokine signaling suppressor 1), and the elevation of DJ-1 after SCI promotes the production of SOCS 1 (Cai et al., 2022). SOCS1 plays a role in the cellular response to oxidative stress by promoting the degradation of activated Rac1 and inhibiting the production of reactive oxygen species (ROS) (Pedrós et al., 2015). Elevated ROS levels serve as a significant trigger for activating the Nod-like receptor protein 3 (NLRP3)inflammasome (Rubartelli, 2012). Recent evidence further underscores the critical role of NLRP3 inflammasome in driving neuroinflammatory processes (Heneka et al., 2013). Following SCI, the expression of the NLRP3 inflammasome is markedly increased (Luo et al., 2019). Upon activation, it controls the maturation and secretion of key pro-inflammatory cytokines, including IL-18, IL-1β, caspase-1 (Juliana et al., 2012). Within the NOD-like receptor (NLR) family, the NLRP3 inflammasome is predominantly recognized as the archetypal member. Its role is well-documented in the development of neurodegenerative and cerebrovascular diseases (Shippy et al., 2020; Xu et al., 2022). After SCI, NLRP 3 inflammasome was also activated, and blocking NLRP 3 inflammasome activation could alleviate the neuroinflammatory response and improve neural recovery (Jiang et al., 2019). MMP-9, a gelatinase predominantly secreted by infiltrating neutrophils, functions as a key mediator during the early inflammatory phase (Lee et al., 2012b). Following SCI, invading neutrophils release MMP-9, an enzyme capable of degrading the extracellular matrix, tight junction proteins, and other peripheral substrates (Zendedel et al., 2018). Furthermore, activation of the NLRP3 inflammasome initiates downstream signaling cascades and promotes the release of inflammatory cytokines, thereby establishing a pro-inflammatory milieu. This environment subsequently influences the expression levels of both MMP-9 and tight junction proteins, exacerbating the disruption of BSCB (Sozen et al., 2009; Xiang et al., 2025). In summary, DJ-1 blocked the activation of NLRP 3 inflammasome by SOCS 1/Rac 1/ROS pathway, thereby alleviating the damage of blood-spinal cord barrier and related neuroinflammation. DJ-1 is a potential target for treating SCI and deserves further exploration (Cai et al., 2022) (Figure 4C).

3.1.6.4 Increased expression of MMP8

MMPs, a group of ECM proteins, are primarily recognized for their capacity to degrade various ECM constituents. Contemporary studies have greatly expanded the catalog of identified MMP substrates. This extended substrate profile includes diverse molecules such as other proteases, protease inhibitors, coagulation factors, chemokines, growth factor precursors along with their binding proteins, cell surface receptors, intercellular adhesion molecules, and virtually all structural constituents of ECM (Agrawal et al., 2008). MMPs are classified as zinc- and calcium-dependent endopeptidases, with a total of 23 distinct members identified within this family (Page-McCaw et al., 2007). The overall understanding of MMPs is that early inhibition of MMPs can maintain BSCB integrity, reduce cell apoptosis, reduce neuroinflammation and provide early and long-term neuroprotection (Zhang et al., 2011). Following SCI, the expression profiles of MMP family members undergo substantial alterations. These enzymes promote the migration of inflammatory cells into the lesion site and are involved in the initial disruption of BSCB. During their transmigration across the vascular wall, leukocytes release MMPs. These enzymes then degrade proteins associated with tight junctions as well as the adjacent basement membrane (Rosenberg and Yang, 2007). We have already described MMP-3 and MMP-9 damage to BSCB. Recent studies have shown that MMP-8 is also related to BSCB disruption and neuroinflammation after SCI. Here are the details:

MMP-8 expression is highest on day 1 (Kumar et al., 2018). Elevated expression of MMP-8 is observed during the early phase (days 1 and 3) following SCI, suggesting a strong association between MMP-8 and the acute inflammatory stage of SCI. Our observations further indicate that peak MMP-8 levels temporally coincide with maximal neutrophil infiltration and the height of inflammatory activity (Furlan et al., 2025). Given that inflammation represents a hallmark feature of SCI, it is notable that the inflammatory response within the injured human spinal cord demonstrates a high degree of similarity to that observed in rodent models (Donnelly and Popovich, 2008; Fleming et al., 2006).

MMP-8 induces endothelial cell injury through the degradation of TJ proteins. The primary roles of these TJs are to establish both a “barrier” and a “fence” function in BSCB, thereby governing its selective permeability (Bazzoni and Dejana, 2004). TJs are essential for BSCB function regulation, and TJ damage increases permeability and affects BSCB function (Bazzoni and Dejana, 2004; Dejana, 2004). MMP-8 upregulates the expression of TNF-α, iNOS, and IL-6 following SCI, while also exacerbating BSCB disruption and cellular injury. One specific mechanism by which MMP-8 impairs the BSCB involves the downregulation of TJ proteins occludin and ZO-1. Structurally, the BSCB is defined by the presence of TJs between adjacent endothelial cells, complemented by a limited degree of transcellular transport. The loss of BSCB integrity and lymphocyte infiltration at the injury site aggravates inflammation, and BSCB rupture after SCI leads to leukocyte infiltration including neutrophils and tissue damage, resulting in secondary injuries including inflammation (Abbott et al., 2006; Bai et al., 2024; Hawkins and Davis, 2005; Kumar et al., 2018; Zlokovic, 2008) (Figure 4D).

3.2 Subacute stage (48 h-14 days)3.2.1 Time: 3 days3.2.1.1 Neutrophil extracellular trap (NETs) production

Neutrophils represent the initial cohort of inflammatory cells that traverse BSCB and infiltrate the site of injury (Aubé et al., 2014; Carlson et al., 1998; Hsu et al., 2006; Mautes et al., 2000), and they are detrimental to SCI (Feng et al., 2021). After the injury, neutrophils rapidly infiltrate the spinal cord within 1 h, reaching the peak level within 24 h (Carlson et al., 1998; Mautes et al., 2000), and persist in the lesion area for 10 days (Mautes et al., 2000). After migrating to the site of injury, neutrophils generate and secrete various pro-inflammatory mediators. These include oxidases (e.g., myeloperoxidase, MPO), proteolytic enzymes (such as matrix metalloproteinase-9 and elastase), and ROS. Collectively, these released factors play a significant role in inducing secondary tissue injury and worsening neurological impairment (Kolaczkowska and Kubes, 2013; Yates et al., 2021). In addition to releasing cytotoxic products, it has been recently discovered that neutrophils cause various diseases by releasing NET [a type of extracellular fibrous network first reported by Brinkmann et al. (2004), Manda-Handzlik and Demkow (2019), and Morishima et al. (2024)].

Following SCI, infiltrating neutrophils generate NETs. NETs formation peaks approximately three days post-injury and subsequently drives neuroinflammatory responses and compromises the integrity of BSCB (Feng et al., 2021). These early pathophysiological alterations are predominantly characterized by local neuroinflammation and BSCB impairment. These two processes engage in reciprocal interactions, ultimately contributing to the exacerbation of secondary injury following SCI (Ahuja et al., 2017; Alizadeh et al., 2019). Preventing neuroinflammation and BSCB damage is a key measure to interrupt persistent secondary damage (Cox et al., 2015; Kumar et al., 2017). Research indicates that diminishing NETs mitigates neuroinflammation and BSCB disruption within the injured spinal cord region, ultimately fostering tissue repair and advancing the recovery of neurological functions (Feng et al., 2021).

The possible mechanism of NETs aggravating BSCB disruption is as follows: After SCI, NETs induce an increase of TRPV4 (transient receptor potential vanilloid type 4) in endothelial cells, and the non-selective cation channel TRPV4 has been shown to promote endothelial damage and BSCB disruption after SCI (Kumar et al., 2020). Cl-amidine functions as a peptidyl arginine deiminase 4 (PAD4) inhibitor, targeting the central enzyme responsible for NETs formation. Both suppression of NETs generation via PAD4 inhibition and breakdown of existing NETs using DNase 1 have been shown to reduce BSCB disruption and inflammatory responses (Jorch and Kubes, 2017; Kang et al., 2020; Vaibhav et al., 2020). Overall, NETs intensify neuroinflammation and BSCB impairment, potentially worsening secondary injury after SCI through upregulation of TRPV4. Conversely, impeding NETs formation or facilitating their clearance mitigates tissue damage and enhances motor recovery. These findings indicate that NETs represent a promising therapeutic target in the management of SCI (Feng et al., 2021; Tang et al., 2024) (Figure 5A).