Spinal cord injury (SCI) is a condition in which traumatic violence to the spine causes spinal fractures and dislocations, resulting in damage to the spinal cord and cauda equina and impairment of motor function, sensory function, nerve reflexes, and sphincter function in the limbs below the plane of injury (1–3). Epidemiological studies over the past 30 years have shown that the incidence of SCI is increasing year by year, with the number of cases reaching 250,000 to 500,000 per year, causing huge losses and heavy burdens to patients, families, and society (1, 4). There are two main stages in the development of SCI: the first stage is primary injury, which is the result of mechanical external forces acting on the spinal cord, mainly in the form of damaged blood vessels, ruptured axons, ruptured nerve cell membranes, etc. (5–11). The second stage belongs to secondary injury, a period of delayed tissue destruction, including vascular dysfunction, edema, ischemia, excitotoxicity, electrolyte transfer, free radical production, inflammation, suppressed apoptotic cell death, etc. (12–14).

Mitochondria are semi-autonomous organelles that not only provide energy for cellular activities but also play a crucial role in the production of reactive oxygen species, programmed cell death, redox signaling, transmembrane transport of various ions and regulation of electrolyte homeostasis (15–18). Copper is among the indispensable heavy metals present in the human body. It is a cofactor for many enzymes in the body and has a strong ability to bind to proteins, as well as having strong redox properties, and it is involved in regulating and maintaining the homeostasis of the body’s internal environment (19). However, excess copper in the body promotes the accumulation of lipidated proteins and the destabilization of iron–sulfur proteins (Fe-S cluster proteins). This leads to proteotoxic stress and, ultimately, cell death; a new mechanism of cell death called cuproptosis is classified as programmed cell death (20–22). Previous studies have demonstrated that copper homeostasis is intricately associated with mitochondrial regulation (23). Copper primarily exists within mitochondria through cytochrome C oxidase (COX) and superoxide dismutase (SOD1), which directly bind to the lipidated components of the tricarboxylic acid cycle (TCA), causing the accumulation and dysregulation of these proteins, obstructing the TCA cycle and initiating proteotoxic stress, among other interventions in various biological processes (21, 24). After the occurrence of SCI, the homeostasis of the body’s internal environment suffered severe trauma, and the metabolism of copper ions was also thrown into disarray, which was inextricably linked to the abnormal fluctuation of copper ions in cuproptosis (22, 25). At the moment of SCI, the blood–brain barrier, a ‘fortress’ guarding the central nervous system, is breached, its integrity is impaired, and its permeability is instantly increased. The copper ion transport channel, which had been precisely regulated, was like a ‘city gate failure,’ and copper ions in the bloodstream took advantage of the chaos to pour into the spinal cord tissue in large quantities. The local concentration of copper ions soared dramatically, forming the copper overload phenomenon. At the same time, SCI-induced inflammatory response, oxidative stress and other ‘secondary disasters’ further interfere with the normal function of intracellular copper ion transport proteins so that the uptake, distribution, and excretion of copper ions in the cell have been ‘out of control’ (26–28). On the contrary, the copper death mechanism, the abnormal increase or decrease of intracellular copper ion concentration is the key factor that triggers this death procedure, while when copper ions are overloaded, the process of copper death is initiated; while when copper ions are deficient, the dysfunction of related proteins involved in copper transport, binding and metabolism will also affect the mitochondrial metabolism, antioxidant defenses, and other key physiological processes, making the cells more sensitive to other stress factors, and indirectly laying the groundwork for cuproptosis. The imbalance of copper ion metabolism after SCI is just like a ‘key’ that precisely opens the ‘door’ to cuproptosis, and the two are intertwined, together exacerbating the pathological process after SCI (28, 29).

Based on the above, we analyzed the differential expression and immune infiltration of 13 CRGs between SCI and non-SCI based on the SCI dataset. Subsequently, unsupervised clustering based on the presence of differentially expressed cuproptosis-related genes (CRGs) was performed to divide the 38 SCI samples into two clusters of molecules associated with copper death and to analyze the expression profile and immune profile between Clusters. The co-expression modules between Cluster C1 and Cluster C2 were identified by WGCNA, and the differential genes within the significant modules were selected for GO and KEGG analysis. We then constructed RF, LASSO, and SVM models based on the differentially expressed CRGs, determined the optimal prediction model and validated the model performance using the Nomogram model, and constructed receiver operating characteristic (ROC) curves and mountain range plots to show the prediction accuracy and expression of the candidate genes. To further validate the accuracy of the candidate genes, we constructed an SCI rat model, and behavioral scores were used to assess the validity of the model. Perfusion-fixed rat spinal cord tissue specimens were observed using electron microscopy, and candidate gene expression in the rat model was verified by qRT-PCR.

SCI is a devastating neurological pathology that causes major motor, sensory, and autonomic dysfunction, and the most serious damage in SCI is to the mitochondria. When mitochondrial damage leads to axonal degeneration and apoptosis (44, 45), it also causes a weakening of oxidative phosphorylation, a significant decrease in the efficiency of releasing energy through oxidation and generating ATP, disruption of Ca²⁺ homeostasis (46), and the release of cytochrome C and other pro-apoptotic proteins from mitochondria into the cytoplasm, inducing apoptosis (47); while in the secondary damage phase of SCI, vasospasm causes a lack of oxygen and energy supply to the damaged parts of the spinal cord In the secondary phase of SCI injury, vasospasm causes a lack of oxygen and energy supply to the damaged area of the spinal cord and disruption of the mitochondrial electron transport chain, resulting in damage to the mitochondria and the release of large amounts of reactive oxygen species, which in turn aggravates oxidative damage to the damaged area of the spinal cord (48, 49). It has likewise been shown that changes in mitochondrial number, localization, abundance, and substrate utilization are closely related to axonal regeneration capacity and prognosis (50–54).

Trace elements are indispensable nutrients in the animal’s body and are required for normal growth, development, and many physiological functions. Deficiencies in the balance of trace elements are closely related to a wide range of diseases and their prognosis, so maintaining normal homeostasis of all trace elements is essential for cell survival and function (55–60). Like many other metallic trace elements, copper plays an important role as one of the essential elements and as a cofactor for essential enzymes necessary for human activity (61). Under normal conditions, intracellular copper ion concentrations are generally maintained at very low levels, but when copper ions exceed the threshold, they become toxic, which in turn leads to cell death (62–64). The copper-triggered cell death that has been studied in recent years is a new model of cell death, with a specific mechanism whereby excess intracellular copper induces the accumulation of specific lipases, resulting in proteotoxic stress and, ultimately, cell death, which is linked to the mitochondrial tricarboxylic acid (TCA) cycle (21, 24).

This study marks the inaugural investigation and analysis of the varied expression of 13 CRGs between SCI patients and the healthy population. The results showed that seven CRGs exhibited distinct expression patterns between SCI patients and the healthy population, and ATP7B, DLD, and SLC31A1 were up-regulated in SCI patients, while DLST, DBT, LIAS, and LIPT1 were down-regulated. This suggests that there may be a close association between SCI and cuproptosis. Subsequently, we conducted an analysis to explore the correlation between the 7 differentially expressed CRGs, aiming to elucidate the relationship between SCI and cuproptosis. We performed unsupervised cluster analysis on 38 SCI samples based on the 7 CRGs and classified them into Cluster C1 and Cluster C2. To further explore the inter-cluster specificity, we explored the expression of the 7 CRGs among Clusters, and the results showed that ATP7B and SLC31A1 in Cluster C1 expression were higher than those of Cluster C2, while LIPT1 expression was lower than that of Cluster C2. The results of ssGSEA analysis indicated that Cluster C2 exhibited elevated levels of T cells CD8, T cells CD4 memory activated, and T cells gamma delta. Cluster C1 displayed increased expression levels in T cells regulatory (Tregs), Macrophages M0, Macrophages M2, Dendritic cells activated, and Neutrophils. Based on the above, Cluster C2 appears to be more strongly associated with the progression of SCI. Next, we performed WGCNA analysis based on Cluster C1 and Cluster C2 and selected the top 3 key modules (MEgreen, MEturquoise, MEred) genes in terms of relevance for GO and KEGG enrichment analysis, and the results showed that the above genes were mainly involved in the Ribosome, HIF-1 signaling pathway, and other pathways. Based on seven CRGs, three machine learning methods (LASSO, RF, SVM-RFE) were used for feature identification, and four candidate genes (DBT, SLC31A1, LIAS, DLST) were obtained after crossover. The ROC curve results indicated that SLC31A1 exhibited the highest diagnostic value (AUC = 0.958) relatively. According to the results of qRT-PCR experiments, the observed expression of the four candidate genes in the SCI model rats was consistent with the above analysis.

DBT constitutes a crucial element within the branched-chain α-keto acid dehydrogenase complex, an intramitochondrial enzyme complex that plays an important role in amino acid metabolism (65–67). DLST is a subunit of the alpha-ketoglutarate dehydrogenase complex, the second oxidative decarboxylation limiting enzyme of the TCA cycle, which acts mainly in the mitochondria and partly in the nucleus (68, 69). In our study, the finding that DLST expression is down-regulated in the disease would indicate that mitochondrial respiratory chain function would be impaired and energy ATP production would be reduced, failing to meet the energy requirements for repair of damaged spinal cord tissue and neuronal survival. Meanwhile, DBT and DLST are enzymes that are lipidated by copper ions, and they are involved in the metabolic complex that regulates the entry of carbon into the TCA cycle (70, 71). The DBT gene expression disorder, likewise, affects the TCA cycle, leading to disruption of the mitochondrial energy metabolism chain and impaired mitochondrial function, exacerbating the energy crisis of cells in the region of spinal cord injury, which is detrimental to the recovery of neurological function. The SLC31A1 gene, also known as copper transporter protein 1, has a high affinity for copper transport in individual cells and is a key gene for copper ion transport, which is used in the study of cancer-related therapies (64). Abnormal expression of SLC31A1 leads to insufficient or excessive levels of copper ions in mitochondria, affecting the activity of copper ion-dependent enzymes, which in turn affects the function of the mitochondrial respiratory chain, resulting in mitochondrial oxidative damage. LIAS belongs to the lipoic acid synthase family and is a highly conserved enzyme found in prokaryotes and eukaryotes. It participates in lipoic acid metabolism and lipid acylation and is an important regulator. LIAS is involved in Fe-S cluster biosynthesis, which is an important component of the mitochondrial respiratory chain complex. Abnormal expression of the LIAS gene leads to impaired synthesis of Fe-S clusters, reduced activity of the respiratory chain complex, impeded electron transfer, reduced efficiency of ATP generation, and generation of a large amount of ROS, which further damages the mitochondrial membrane and the internal structure of the mitochondrion, forming a vicious circle. FDX1 is an upstream regulator of protein lipidation and is also a direct target of copper ion carriers, which has a role in regulating protein-lipid acylation (72, 73). In previous studies, it was found that deletion of FDX1 and LIAS conferred resistance to copper-induced cell death (21), but it has also been shown that defects in neonatal LIAS are associated with severe metabolic disorders (74). Based on the above mechanism, it is not difficult to judge that the abnormal expression of the above genes also causes oxidative stress and immune response (75, 76). When oxidative stress is combined with mitochondrial injury, DAMPs will be released to activate macrophages, T cells and other immune cells, and the abnormal expression of the genes will indirectly regulate the expression of chemokines and cytokines of immune cells by influencing the function of mitochondria and oxidative stress, and affect the infiltration of immune cells into the site of SCI (77, 78). Normal immune cell function depends on mitochondrial energy supply and redox homeostasis, and mitochondrial damage and oxidative stress caused by genetic abnormalities affect immune cell activity. For example, macrophage polarization is disturbed, which prevents effective immunomodulation, and T-cell activation and proliferation are inhibited, affecting the body’s immune repair ability. Meanwhile, oxidative stress induces the release of inflammatory factors such as TNF-α and IL-1β, etc. Oxidative stress caused by abnormal gene expression promotes the expression and release of inflammatory factors by activating inflammatory signaling pathways such as NF-κB, which further recruits immune cells and exacerbates the inflammatory response, and the excessive inflammatory response will also lead to the dysfunction of the immune cells, resulting in a vicious circle and exacerbating the pathological process of SCI (79–81).

In summary, the four CRG genes, namely DLST, DBT, LIAS, and SLC31A1, obtained from this study based on the crossover of the three models of LASSO, RF, and SVM-RFE had satisfactory results in assessing the immune infiltration and subtype and pathology of SCI.

However, this study still has shortcomings. First, this study obtained a certain number of SRGs from the CellAge database, but more SRGs currently exist that have not been identified. Secondly, the sample size of the three training sets in this study was small, although we compensated by choosing external validation and experimental validation to confirm the candidate genes obtained in this study. Thirdly, although the four candidate genes eventually obtained in this study are closely associated with cuproptosis and SCI, validation by qRT-PCR alone is not sufficient, and it is expected that more in vitro and in vivo studies will be conducted in subsequent research work to confirm the above findings. Future research should focus on exploring the regulatory mechanisms of the candidate genes and developing effective interventions. Multidisciplinary cooperation and big data analysis are expected to provide a solid theoretical foundation and practical guidance for addressing the clinical challenges of related diseases.