Central pain
CP, also referred to as dysesthetic pain syndrome or central dysesthesia syndrome, is a stubborn and common complication of SCI, which may develop spontaneously or may be induced by skin irritation [
16,
17]. The incidence of CP was reported to be 11–94% in 2 million patients with SCI, and about one third have severe CP. Patients may experience a variety of symptoms, including knife-like pain, burning pain, tingling, radiating pain, tight pain, and cold feeling, which may seriously affect their quality of life [
18]. Common treatments for CP after SCI currently include physical therapy, drugs, and surgery. However, the efficacies of these treatments are unsatisfactory, and drug treatment can lead to depression, drug addiction, and other complications [
19]. It is therefore essential to develop a better understanding of the pathogenesis of the disease in order to improve its treatment.
There are several hypotheses regarding the mechanisms of SCI-associated CP, including imbalance of sensory pathways [
20], imbalance of inhibitory and excitatory receptors [
21,
22], problems with pattern-generating mechanisms [
23], and neuroimmunological mechanisms [
24]; however, these hypotheses cannot fully explain the pathogenesis of CP. Recent studies have indicated that certain brain regions, such as the somatosensory center, thalamus, and limbic system, undergo remodeling after SCI to compensate for the loss of sensory function below the injured segment, and many researchers believe that this remodeling may be a key factor in SCI-associated CP [
25,
26].
Patients with spinal cord transection injuries were shown to experience pain in the region distal to the injured segment, while the severity of pain was not determined by the extent of the injury [
27,
28]. We therefore aimed to seek the source of pain proximal to the injured segment.
CP includes spontaneous pain caused by injury itself, with no external stimulus, as well as pain induced by external stimuli. In this study, we established a rat SCI model using the Allen impact method [
29] and found that all rats, except those in the sham operation group, demonstrated excessive grooming behaviors, such as self-biting and scratching the hind limbs, which may have been caused by spontaneous pain in the hind limbs. We also showed that the thermal pain threshold in the lower extremities was significantly decreased in the SCI/CP groups, thus triggering hyperalgesia. These results indicated that both spontaneous and induced pain were induced in this rat model, in line with previous studies [
30‐
32]. Treatment with DFO or arginine significantly increased the thermal pain threshold and reduced excessive grooming behavior, suggesting that these drugs relieved the induced and spontaneous SCI-associated CP.
Iron overload in the brain
Iron is an important trace element in the human body and is widely distributed in the brain, where it participates in many important physiological and biochemical processes, including DNA, RNA, and protein synthesis, myelin synthesis, myelinogenesis, and development, as well as the synthesis of some neurotransmitters such as dopamine. However, excess iron is toxic to the human body, and abnormal accumulation of iron in various brain areas has been found in many neurodegenerative disorders [
33,
34]. Iron is known to induce the production of hydroxyl radicals through the Fenton reaction [
35] (Fe2++ H2O2 → Fe3 + ·OH + OH), thus exacerbating oxidative stress and leading to tissue and cell damage.
Iron is redistributed and deposited in certain brain areas in some neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, and Hallervorde–Spatz syndrome [
36‐
38], and iron levels have also been shown to be increased in brain areas in animals under stress conditions, such as heat stress, exercise stress, and seasickness [
39,
40]. Although increased iron levels have not been conclusively identified as either the cause or consequence of these diseases, recent studies of genes associated with iron metabolism in the brain suggest that increased iron is an initiating factor of neuronal death in some neurodegenerative diseases [
41].
In the present study, we measured iron levels in whole brain and specific brain areas of rats by atomic absorption spectrophotometry and found no significant differences in whole-brain iron levels among the control, SCI, and SCI +
l-arginine- or SCI + DFO-treated rats. However, iron levels in the thalamus, hippocampus, and hind limb sensory area were significantly increased in all SCI/CP groups, with an iron-deposition distribution similar to that in some neurodegenerative diseases. The hind limb sensory area is the projection area of the hind limb in the cortex, while the thalamus and hippocampus are transit areas of sensory transduction. Lesions in these areas can cause CP [
42,
43]. In the current study, iron levels in certain brain areas remained elevated 12 weeks after surgery to induce SCI, supporting long-term deposition of iron in the brain consistent with the long duration of CP.
The fact that iron levels increased in certain brain areas but not in the brain as a whole may be attributable to the function of the blood–brain barrier, which isolates the brain from the external environment. Although iron levels increased in some areas, they remained stable overall, consistent with the results of previous studies [
44].
Mechanisms of iron overload in the brain
Maintenance of iron homeostasis in the brain depends on the normal expression and coordination of a variety of brain iron-metabolism proteins. Penetration of iron through the blood–brain barrier and the uptake of iron by neurons are mainly mediated by the classic Tf/TfR pathway [
45‐
48].We found that TfR1 mRNA and protein levels were increased in the thalamus, hippocampus, and hind limb sensory areas of SCI/CP rats compared with the sham operation group, suggesting that iron overload in these brain areas may be caused by increased iron levels mediated by the Tf/TfR pathway.
Fn is a natural iron chelator that is widely expressed in neurons and glial cells in human and rodent animal brains. Iron stored by Fn accounts for 75% of total brain iron [
49]. The structure of Fn includes H and L subunits, of which the L subunit is associated with long-term iron storage [
50,
51]. Ferrous iron taken up by tissue cells is oxidized to ferric iron and sequestered by Fn and stored in a non-toxic form. [
52] Previous studies have shown that upregulation of Fn may limit iron-induced brain injury [
53]. Furthermore, Fn was found not to increase in line with increasing iron levels in Parkinson’s disease and Alzheimer’s disease [
54], suggesting that free iron may not be captured and stored by Fn, and peroxidation caused by free-iron toxicity may not be prevented.
In this study, levels of the Fn in the thalamus, hippocampus, and hind limb sensory area were significantly decreased in SCI/CP rats compared with the sham operation group, suggesting that the iron-storage capacity was reduced in these areas, potentially leading to increased free iron.
Tfr and Fn play a crucial role in maintaining iron homeostasis in the brain. Expression levels of these two proteins are mainly regulated by the iron-responsive element (IRE)/IRPs. In the absence of iron, IRPs bind to IREs, which upregulate TfR1 synthesis and downregulate Fn synthesis. Iron deficiency thus leads to increased iron uptake and reduced iron storage. In contrast, when iron is in excess, IRPs do not bind to IREs, TfR1 synthesis is downregulated, and Fn synthesis is upregulated, leading to decreased iron intake and increased iron storage. However, IRP activity is not only regulated by intracellular iron content. Previous studies found that the IRE/IRP complex was more stable in the brain of patients with Alzheimer’s disease. Increased stability of this complex can stabilize Tfr mRNA levels and decrease Fn synthesis, thereby increasing iron uptake and decreasing iron storage in the brain in these patients, resulting in brain iron overload and the induction of oxidative stress and neuronal apoptosis.
The results of the current study showed that IRP1 protein levels were significantly upregulated in the hind limb sensory area, thalamus, and hippocampus of SCI/CP rats, compared with the sham operation group, while TfR was increased and Fn was decreased in these areas. This suggests that iron regulation was disrupted in these areas after SCI, and that IRP1 may be the initiating factor for this process. In addition to the Tf/TfR pathway, human Lf/Lf receptor may also play a role in iron transport through the blood–brain barrier. However, we found no differences in Lf expression in the thalamus, hippocampus, and hind limb sensory area among the different rat groups, suggesting that Lf-mediated iron uptake may not be involved in iron overload after SCI.
DFO is a potent iron ion chelating agent, which has been shown to penetrate the blood–brain barrier rapidly and accumulate in the brain parenchyma. It can chelate free iron ions to form a relatively stable compound and effectively prevent the release of iron ions from Fn, thus significantly reducing peroxidation damage caused by iron overload [
23]. In our study, DFO significantly reduced iron levels in the hippocampus, hind limb sensory cortex, and thalamus, and increased Fn levels in these areas, suggesting that DFO may chelate the iron in these areas and inhibit the degradation of Fn. However, whole-brain iron levels were not reduced, suggesting that DFO decreased iron levels by binding iron ions and inhibiting the formation of free iron, rather than by promoting the excretion of iron outside the brain.
Iron overload and oxidative stress
There are two types of free radical scavengers in the human body, antioxidases and antioxidants, which provide electrons to reduce and block the formation of oxygen free radicals and thus prevent cell damage. Free radicals can induce scavenger enzymes to maintain a dynamic balance [
55]. The central nervous system is rich in lipids, and free radicals can cause lipid peroxidation and induce pathological changes in cell morphology and function. Arachidonic acid substances generated by lipid peroxidation, including prostaglandins D2, E2, and I2 and leukotrienes, can act on the nerve and glial cells to injure cell membranes and cause cell dysfunction [
56].
SOD and MDA reflect the degree of lipid peroxidation. MDA is a metabolite of the peroxidation reaction of membrane unsaturated fatty acids induced by oxygen free radicals, and an indicator of the degree of cell damage [
57]. SOD is a natural antioxidase capable of cleaning oxygen radicals, thereby blocking the lipid peroxidation chain reaction, and providing an indicator of the body’s ability to scavenge oxygen free radicals. Previous studies found that iron overload increased serum and organ levels of MDA and reduced SOD [
58,
59]. Given that increased brain iron levels and accompanying oxidative stress have been identified in many neurodegenerative diseases, some researchers have suggested that abnormal increases in brain iron may lead to the generation of large numbers of free radicals and further induce cell death, which may be one reason for the observed neuronal death in neurodegenerative diseases. The addition of ferrous sulfate in the diet has been shown to increase iron ion levels and oxygen free radicals in rat brains, and to cause neuronal injury, and even death [
12]. Following cerebral hemorrhage and cerebral ischemia reperfusion, excessive iron ions can catalyze lipid peroxidation to produce oxygen free radicals, which attack cell proteins and nucleic acids leading to peroxidation injury, which represents an important mechanism of brain injury secondary to cerebral hemorrhage [
60‐
64]. The potential causative role of oxidative stress caused by iron overload in the brain in neurodegenerative diseases is supported by the abnormal accumulation of iron in certain brain areas in patients with Parkinson’s disease and Alzheimer’s disease [
65].
In the present study, SOD activity was significantly decreased and MDA content was significantly increased in brain tissues of rats with CP. This suggests that large numbers of oxygen free radicals were produced in the brain, and that decreased SOD activity caused lipid peroxidation injury of biomembranes, which increased the content of MDA. Intervention with DFO decreased iron levels and restored SOD activity, associated with decreased MDA content.