Therapeutic targets and nanomaterial-based therapies for mitigation of secondary injury after spinal cord injury

Abstract

Spinal cord injury (SCI) and the resulting neurological trauma commonly result in complete or incomplete neurological dysfunction and there are few effective treatments for primary SCI. However, the following secondary SCI, including the changes of microvasculature, inflammatory response and oxidative stress around the injury site, may provide promising therapeutic targets. The advances of nanomaterials hold promise for delivering therapeutics to alleviate secondary SCI and promote functional recovery. In this review, we highlight recent achievements of nanomaterial-based therapy, specifically targeting blood–spinal cord barrier disruption, mitigation of the inflammatory response and lightening of oxidative stress after spinal cord injury.

Spinal cord injury (SCI) and the resulting neurological trauma commonly result in complete or incomplete paralysis, hypoesthesia, autonomic disorder and other sequelae. Although the incidence rate of acute SCI has remained stable in the USA, the annual cases continued to increase as the national population increased [1]. For a patient with cervical trauma, the medical expenses in the first year after the injury were about US$1 million in 2013 [2]. Due to the decline of individual independent living ability, the estimated lifetime economic burden per individual with this condition is more substantial [3,4]. In addition, problematic secondary health conditions related to SCI, such as bladder/bowel control and pain, significantly impact patients’ social participation [5]. Therefore the treatment of SCI is important for patients in terms of both medical and social problems.

Various trauma, including traffic accidents and sports or labor injuries, can destroy the stability and integrity of the spinal anatomical structure, causing traction, extrusion, rotation, contusion and tearing of the spinal cord and direct death of neurons, which is the primary injury of SCI. Because the occurrence of SCI is unpredictable, primary SCI can only be mitigated by minimizing its severity. Current decompression surgery can relieve mechanical pressure and restore the spinal stability, while systemic administration of high-dose methylprednisolone (MP) may mitigate inflammation and parenchymal edema. Even though both these regimens partially preserve neural tissue, other strategies are needed to repair the damaged tissue and promote functional recovery after SCI.

Secondary injury after SCI is the subsequent pathophysiological changes caused by the primary insult, including changes of microvasculature and the inflammatory response around the injury site. These changes are defensive responses to trauma, aiming to limit insult and maintain microenvironmental balance; however, they may also aggravate the damage, enlarge the lesion volume and arrest functional recovery. Compared with the primary injury, secondary injury progressively evolves over days to weeks [6]. It is a complex, multifaceted process and therefore offers a window of opportunity for therapeutic intervention at acute and chronic time points post SCI. Among various potential mechanisms within the secondary injury cascade, we will focus on the strategies for prevention of blood–spinal cord barrier (BSCB) disruption, mitigation of the inflammatory response and alleviation of oxidative stress.

One strategy to mitigate secondary injury progression following SCI and promote neuroprotection, neurorepair and neuroregeneration is to utilize nanomaterial-based delivery systems to deliver single or multiple drugs to treat these complex pathophysiological events. Nanomaterials and their polymeric components are attractive delivery systems for the spinal cord because of their potential to cross the BSCB, delivery efficiency and targeting capability. Furthermore, some components of nanoparticles, such as polyethylene glycol (PEG), can themselves slow the progression of secondary damage by several mechanisms [7,8]. In this review we will focus on therapeutic strategies for the prevention of BSCB disruption, mitigation of inflammatory response and alleviation of oxidative stress based on nanomaterial-based therapy targeting the secondary injury (Figure 1). Table 1 shows a summary of the nanomaterial-based therapy discussed in this review. Animal models used to study pathophysiology, mechanism of injury and investigate therapeutic strategies to improve functional recovery in SCI are discussed.

Figure 1. . Spinal cord injury pathology and nanomaterial-based therapies.

Table 1. . Summary of nanomaterial-based therapy for inhibiting secondary injury after spinal cord injury.

Pathology	Nanomaterials	Therapeutic targets	Ref.
BSCB disruption	Titanium nanowire	Modulating activity of lipoxygenase and cyclooxygenase	[9,10]
	TiO2 nanowire	Reducing free radical and neurotoxic agents	[11]
	NBP nanowire	Inhibiting ER stress	[12,13]
	Solid lipid nanoparticle	Inhibiting ER stress	[14]
	Biodegradable PLGA nanoparticles	Reducing TNF-α/ROS	[15–17]
Neuroinflammation	CMCht/PAMAM dendrimer nanoparticles	Reducing TNF-α/ROS	[18]
	PLGA nanoemulsion	Reducing TNF-α/ROS	[19]
	Albumin nanoparticles	Reducing TNF-α/ROS	[20]
	PLGA nanoparticles	Inhibiting CDK-1	[21]
	DS-MH chelate complex	proNGF production in microglia	[22,23]
	PgP polymeric micelle nanoparticles	Inhibiting PDE-4 and restoring cAMP levels	[24]
	PLGA-IMPs	Reducing M1 macrophage polarization and microglia activation	[25]
	PCL nanoparticles	Modulating the resident microglial cells	[26]
	MoS2 PEG nanoflowers	Inhibiting the expression of TNF-α, CD86 and iNOS; promoting the expression of Agr1, CD206 and IL-10	[27]
	PMMA nanoparticles	Modulating activated macrophages and microglia	[28]
	Lipidoid nanoparticles	Facilitating M1-to-M2 transition	[29]
	PEG (MW2000)	PEG-mediated membrane repair	[30]
Oxidative stress	PEG-modified silica nanoparticles	Reducing the formation of ROS and the process of lipid peroxidation of the membrane	[31]
	PEG-coated nanozyme	Scavenging ROS	[32]
	MPEG–PLLA–PTMC copolymer	Modulating cytosolic Ca2+ concentrations	[33]
	Iron oxide nanoparticles	Attenuating free radicals	[34]
	Mesoporous silica nanoparticles	Acrolein trapping	[35]
	Chitosan nanoparticles	Acrolein trapping	[36]

BSCB: Blood–spinal cord barrier; CMCht/PAMAM: Carboxymethylchitosan/polyamidoamine; DS-MH: Dextran sulfate-minocycline hydrochloride; ER: Endoplasmic reticulum; MoS2 PEG: Molybdenum disulfide polyethylene glycol; MPEG-PLLA-PTMC: Monomethyl poly(ethylene glycol)-poly(l-lactide)-poly(trimethylene carbonate); NBP: Nanowired Dl-3-n-butylphthalide; PCL: Poly-ε-caprolactone; PDE-4: Phosphodiesterase 4; PEG: Polyethylene glycol; PgP: PLGA-graft-polyethylenimine; PLGA: Poly(lactic-co-glycolic acid); PLGA-IMP: Poly(lactic-co-glycolic acid)-immune-modifying microparticle; PMMA: Poly(methyl methacrylate); ProNGF: Pro-nerve growth factor; ROS: Reactive oxygen species; TiO₂: Hydrogen titanate.

Hyperpermeability of the blood–spinal cord barrier

After a primary insult occurs, the spinal intraparenchymal blood supply, namely spinal microcirculation, will change and cannot be corrected by surgery. Along with the microhemorrhages that occur immediately and ischemia that develops progressively, the permeability of vessels change, leading to interstitial edema of the neural tissue and dysfunction of ascending/descending axonal fibers. In the spinal cord tissue, there is a similar structure to the blood–brain barrier, the BSCB, which can regulate the efflux of components from the endovascular cavity and maintain homeostasis in the neural parenchyma. As shown in Figure 2, the BSCB is composed of glial cells, pericytes, endothelial cells and tight junctions (TJs), which consist of claudins, occludin, zonula occludens and junctional adhesion molecules. Following traumatic injury, the structure of the BSCB is disrupted and its permeability is increased, which results in infiltration of immune cells and inflammatory factors entering the stroma, ultimately causing spinal cord edema and increasing intramedullary pressure [37–39]. A study about the spatial and temporal vascular changes in a rat SCI model showed maximal BSCB disruption at 24 h post injury, with significant disruption observed up to 5 days post injury [40]. Moreover, dynamic contrast-enhanced MRI studies demonstrated the BSCB remained compromised at 56 days after SCI and that there was a significant link between the neurobehavioral function and the restoration of BSCB integrity [41].

Figure 2. . Blood–spinal cord barrier disruption after spinal cord injury.

BSCB: Blood–spinal cord barrier.

Several proteases, such as angiopoietins and matrix metalloproteinases (MMPs) regulate the integrity of BSCB. Ang1 has been verified to regulate permeability in cultured rat spinal cord microvascular endothelial cells [42]. In mouse SCI models, there was a rapid reduction of Ang1 mRNA levels at 6 h post injury. Administration of a miR-711 hairpin inhibitor after SCI significantly rescued Ang1 expression associated with the protection of epicenter blood vessels [43]. In mouse models of contusion SCI, daily intravenous injections of an Ang1 mimetic significantly reduced the amount of luciferase that extravasated into parenchyma (61 ± 13% vs vehicle group [100 ± 17%] at 72 h; p < 0.05), which suggests that Ang1 treatment reduces pathological permeability [44]. Schimke et al. combined Ang1 with nanoscaled diamond particles and implanted them into osseous defects in sheep calvaria, which demonstrated that the combination enhanced vessel growth [45]. In a moderate-compression SCI model, Mmp8 expression significantly increased and administration of an MMP8 inhibitor reduced the amount of Evans blue dye extravasation after SCI, which suggests inhibiting Mmp8 may attenuate SCI-induced BSCB breakdown [46]. Moreover, compared with wild-type SCI mice, BSCB permeability is significantly lower in Mmp8, Mmp9 or Mmp12 knockout SCI mice and this coincides with improvements in the locomotor recovery [38,47,48]. In another study, fluoxetine treatment was shown to significantly prevent BSCB breakdown and improve locomotor function after injury by inhibiting the expression of Mmp2, Mmp9 and Mmp12 [49]. Similarly, Jmjd3 mRNA depletion can simultaneously inhibit Mmp3 and Mmp9 expressions and significantly attenuate BSCB permeability [50]. Together, these data demonstrate that both Ang1 and MMPs are critical mediators of BSCB permeability after SCI.

TJs play a major role in regulating the selective permeability of the endothelium by physiologically controlling the paracellular diffusion of compounds from blood [51]. Different regulatory factors may change the permeability through affecting the expression of TJs [52–54]. In miRNA-125a-5p-overexpressing cells, the expressions of ZO-1, occludin and their mRNA significantly increased, indicating lower permeability compared with the control. By contrast, the expression of the TJ components was significantly reduced in miRNA-125a-5p-inhibited cells, coinciding with higher cell membrane permeability. The study demonstrated that miRNA-125a-5p can reduce the permeability of the BSCB by upregulating the expression of TJs [52]. In a study related to the neuroprotective effects of propofol on SCI, intravenously infused propofol suppressed the decrease in expression of occludin and claudin-5, attenuated the increase in BSCB permeability and decreased histological damage [53]. Moreover, a novel anti-inflammatory and antioxidative chemical, HO-1 fragments lacking 23 amino acids at the C-terminus (HO-1C[INCREMENT]23), was delivered by an adenoviral carrier into an SCI rat model and showed smaller reductions in TJ proteins and capillary permeability as well as better kinematic functional recovery than the control groups [54].

Another strategy to reduce secondary damage after an SCI may be restoring the BSCB integrity and reducing vascular permeability. Topical application of neurotrophins, such as growth hormones, showed that it can reduce the early manifestations of microvascular permeability disturbances and edema formation following trauma [55]. Several other studies have shown that the goal of repairing SCI can be partially achieved by mitigating changes in vascular permeability [56–58].

Some compounds, such as AP-173, AP-713 and AP-364, were delivered TiO2 based nanowires (hydrogen titanate) and evaluated the therapeutic efficacy of nanowired drugs on the breakdown of BSCB and sensory motor function after an SCI [9,10]. The nanowire is a kind of crystalline ceramic biomaterial based on TiO₂ (hydrogen titanate). Compared with the normal compound, nanowired AP-713 compound significantly reduced BSCB disruption and attenuated behavioral dysfunction [9]. Because nanowires alone have no therapeutic effect, it is speculated that the nanowire can enhance the compound’s inherent neuroprotective effects. The temporal efficacy of nanowired AP-713 treatment was studied by comparing a delayed topical application (60 min after injury) with rapid administration (5 min after injury) [10]. Interestingly, the delayed application of nanowired AP-713 still had obvious therapeutic effects, whereas delayed administration of normal compound AP713 showed significantly reduced efficacy and beneficial effects compared with delayed or rapid intervention. This suggested that nanowired compounds have prolonged neuroprotective effects in SCI even when applied at a subacute time frame. Although the study did not identify the neuroprotective mechanism of nanowire, the possible reasons could be increased bioavailability and reduced biodegradation.

TiO₂-nanowired growth hormone has been shown to induce longer neuroprotection after a focal SCI in rats by immediate intrathecal administration; it could significantly reduce BSCB disruption, edema formation and neuronal injuries after trauma up to 24 h after injury, whereas normal growth hormone given under identical conditions alleviated pathology only when administered after no longer than 8 h [11]. Another series of experiments explored the use of nanowired Dl-3-n-butylphthalide (NBP). When nanowired NBP was immediately applied intravenously or topically after SCI, it significantly enhanced the reduction of the BSCB breakdown and edema formation compared with NBP alone [12,13]. Because all those studies only performed short-term functional studies (no more than 24 h after injury), further studies are needed to investigate more continuous functional improvements to clarify the effectiveness of the nanowire drug-loading system.

Besides the nanowire, the solid lipid nanoparticle drug-delivery system has also been investigated in the context of repairing BSCB [14]. Although carbon monoxide-releasing molecules (CORMs) play a protective role for vessels, they have low solubility and a very short half-life (~1 min) [59]. In response to those limitations, CORM-2 was incorporated into a solid lipid nanoparticle (CORM-2-SLN). In vitro assays showed that CORM-2-SLN constantly and slowly released CO, resulting in a 50-times elongation of CO release half-life compared with that of CORM-2 in solution (50 vs 1 min half-life). In a T10 compression SCI rat model, the efficacy of CORM-2-SLN was significantly better than that of CORM-2 in solution, as demonstrated by protection of endothelial cells, inhibition of apoptosis, prevention of BSCB disruption, reduction of lesion volume and improvement of neurological outcomes. Although the exact mechanism needs to be elucidated, the superior results of CORM-2-SLN may be attributed to the constant and slow CO release by SLN.

Inflammatory response

After the spinal cord is injured, not only will the morphological structure change, but also further complex pathophysiological changes will occur due to the response of various cells and the effector molecules they secrete. As shown in Figure 3A, the resting microglia will be activated in the spinal cord as the first responders due to the primary damage. Within minutes of injury, microglial cells can transform into an ameboid morphology and migrate to the injury site [60]. Accompanying increased vascular permeability, blood-derived immune cells such as neutrophils and monocytes will infiltrate the spinal cord parenchyma and exacerbate the inflammatory response. In rats, neutrophils in the parenchyma peaked in number at 24 h after SCI, while in humans neutrophils increased 15-fold in the injured spinal cord tissue 1–3 days following injury [61,62]. In both rats and humans the blood-borne monocyte/macrophages infiltrated the lesion later and were present for weeks to months after the SCI [61,62]. The role of these inflammatory cells is controversial. Several studies have shown that reducing infiltration of neutrophils or inhibiting their function can significantly improve the prognosis of an SCI [63–65]. However, another study has shown that when neutrophils were reduced 6–12 h post injury, dysfunction was aggravated, suggesting that neutrophils are beneficial for repair [66]. Similarly, macrophages also have been shown to elicit diverse functions, which will be discussed in detail later. TNF-α and IL-1β have long been considered as important cytokines potentiating the secondary injury response following an SCI [6,67–70]. However, a recent study of cytokine profile at different periods after contusion injury showed no significant change in concentrations of either TNF-α or IL-1β in the site of SCI, though there were alterations in upstream mediators of both cytokines [71]. Based on the knowledge of inflammation’s involved in spinal cord secondary injury, anti-inflammatory treatment will be the one of the therapeutic options.

Figure 3. . Inflammatory response after spinal cord injury.

Methylprednisolone

The steroidal drug MP is an FDA-approved synthetic glucocorticoid which elicits an anti-inflammatory effect in many pathological conditions. Bracken et al. first reported that systemically administered MP was effective within 8 h following an acute SCI [72]. However, Hugenholtz et al., suggested that there was insufficient evidence to support the use of high-dose MP, based on their systematic analysis of the existing clinical Level I and II evidence pertaining to MP infusion following acute SCI [73]. Although systemic administration of MP is still an option for acute human SCI, the use of high-dose steroids has significantly decreased due to side effects such as pneumonia, gastrointestinal tract hemorrhage, urinary tract infections and elongation of average hospitalization [74,75]. Thus a treatment strategy of local delivery of MP to reduce these systemic side effects is necessary.

Several nanocarrier strategies have been developed to deliver MP locally in a selective and sustained/controlled manner to modulate inflammatory responses. Biodegradable poly(lactic-co-glycolic acid) (PLGA) nanoparticles (NP) have been used to deliver MP locally in the injured spinal cord [15]. In a T9–10 rat over-hemisection model, local MP-NP injection significantly reduced the number of ED-1⁺ macrophages/reactive microglia, decreased lesion volume and improved functional outcomes of the injured SC compared with three other groups (local non-MP nanoparticles administration, local normal MP injection and systemic administration of MP). Most importantly, the encapsulation of MP by PLGA administered near the lesion site contributed to a dramatically lower dose of MP (400 μg/animal) required to produce both cellular and functional improvements, which was approximately 5% of the systemically administered dose (7∼8 mg/animal). Aimed at eliminating side effects resulting from systemically administered MP, other similar MP nano-carriers have recently been developed to elevate plasma concentration and increase bioavailability [16,17].

MP was also delivered by carboxymethylchitosan/polyamidoamine (CMCht/PAMAM) dendrimer nanoparticles intracellularly to microglia, which elicit the first inflammatory response [18]. In an in vitro assay, sustained (14 days) release of MP by the nanoparticles was longer than that of another MP-encapsulating PLGA nanoparticle, which was reported to slowly release MP over 4 days [15]. The intracellular internalization of MP-loaded CMCht/PAMAM dendrimer nanoparticles modulated the metabolic activity of microglia cells in cell culture. At 4 weeks post-hemisection injury, SCI rats treated with MP-loaded nanoparticles recovered a weight-supporting stance of their hind limbs in the open field, whereas SCI rats in the saline-treated, nanoparticle vehicle-treated and MP solution-treated groups were only able to generate slight isolated joint movements, suggesting that recovery of locomotor function was significantly improved with MP-NP treatment. While this study showed promising behavioral recovery, it lacked relevant histological evidence and the related changes of inflammatory cytokines.

One of the advantages of nanomaterial delivery systems can be modification of surface moieties by conjugating to achieve targeted therapy. Given the upregulated expression of the albumin receptor albondin (a 60-kDa glycoprotein) at the SCI site [19], albumin can be the most suitable targeting moiety. Chen et al. evaluated the efficacy of MP delivered by an albumin-conjugated chitosan-stabilized, lecithin-emulsified, multilayered nanoemulsion (albumin-MNE) [76]. In both the beam walk test and the grid walk test, albumin-MNE-treated SCI rats performed significantly better than those treated with non-decorated MNE. In histological analysis, albumin-MNE significantly reduced the Bax-mediated mitochondrial apoptosis at SCI compared with the non-decorated MNE group. The data showed that albumin-MNE treatment improved histopathological conditions and behavior of SCI rats in a more effective manner than treatment with non-decorated MNE.

In the similar nano-drug loading system, both MP and minocycline were delivered together [19]. In the treatment of hemisectioned SCI, coadministration of MP and minocycline in an albumin-conjugated nanoemulsion significantly reduced lesion volume and improved behavioral outcomes when compared with a nonconjugated nanoemulsion. The results suggested that site-specific controlled release of therapeutics is important for the treatment of SCI.

In another study, albumin itself was also used as a nanocarrier for MP. Lin et al. developed MP-loaded albumin nanoparticles decorated with NEP_1–40, a Nogo receptor antagonist, to target Nogo receptors on the neuronal cell surface. They reported that a single intrathecal injection of NEP_1–40-MP-NPs reduced infiltration of immune cells and improved motor function compared with the other control groups (saline, blank NPs, MP or MP-NPs) at 28 days post injury [20].

Other anti-inflammatory drugs

The cell-cycle inhibitor flavopiridol has been shown to improve functional recovery in SCI animal models, although the systemic dose of flavopiridol may cause secretory diarrhea [77]. Flavopiridol-encapsulated PLGA nanoparticles (flavopiridol-NPs) were topically delivered in the lesion site in a rat hemisection SCI model. In the flavopiridol-NP-treated group, cell-cycle activation, glial scarring and expression of inflammatory cytokines were significantly decreased and neuronal survival and regeneration were facilitated compared with the group treated with blank NPs. The cavitation volume in the flavopiridol-NP-treated group was decreased by 90% compared with the blank-NPs-treated group. Meanwhile, the motor function recovery in the flavopiridol-NP-treated group had significantly improved at 6 weeks post injury compared with the blank-NPs treated group [21].

Minocycline hydrochloride (MH) is an antibiotic and anti-inflammatory drug and has been shown to treat secondary injury after traumatic SCI [78–81]. However, the systemic doses used in these studies are too high to be used in human patients due to hepatotoxicity. To avoid the hepatotoxicity induced by systemic high doses of MH, Zhang et al. developed dextran sulfate–MH (DS-MH) chelate complexes using divalent metal ions such as Ca²⁺ and Mg²⁺, ranging from 50 to 100 nm, for stable long-term MH release [22]. They further developed the formulation, which can release a high dose of MH for the acute stage and a low dose of MH for the chronic SCI treatment, using agarose hydrogel. The DS-MH complexes were designed to burst-release early followed by later slow release, to achieve high doses of MH at the acute stage for neuroprotection and low doses of MH at the chronic stage targeting chronic inflammation and minimizing the potential toxicity. The DS-MH complexes were embedded in injectable agarose hydrogel so that they remained localized at the injury site after the gel was injected into the intrathecal space. In a unilateral C5-level contusion injury rat model, local delivery of DS-MH complexes at a dose lower than the standard human dose (3 mg/kg) was more effective in reducing secondary injury and promoting locomotor functional recovery than systemic injection of MH alone (90–135 mg/kg) [23].

Rolipram (Rm), a PDE4 inhibitor, can restore decreased levels of intracellular cAMP in the spinal cord tissue after an SCI [82]. Although several studies have suggested that neuroprotection of Rm contributes to its potential of promoting axonal regeneration, in one study Rm-loaded nanoparticles were also demonstrated to significantly mitigate the inflammation [24]. Macks et al. developed a cationic amphiphilic copolymer, PLGA-graft-polyethylenimine (PgP) as a drug and therapeutic nucleic acid carrier; they reported that water solubility of Rm was increased up to ∼6.8-times by dissolving Rm in the hydrophobic core of the PgP micelle. They observed that administering Rm-PgP (10 μg Rm) via intraspinal injection in the spinal cord lesion site restored cAMP levels to sham animal levels, reduced apoptosis and the inflammatory response, and increased neuronal survival in the rat SCI compression injury model [24].

Immunomodulation of microglia/macrophage by nanomaterials

As mentioned above, studies have demonstrated that there are two different macrophage phenotypes in the inflammatory process (Figure 3B). The classically activated macrophage phenotype, M1, is activated by IFN-γ and secretes proinflammatory cytokines such as TNF-α and IL-1β. A second phenotype, alternatively activated macrophage or M2, is activated by IL-13 or IL-4 and reduces inflammation and promotes tissue repair. Similar to infiltrating monocytes/macrophages post-SCI, activated microglia at SCI lesion sites may also acquire a M1 or M2 polarized phenotype with divergent effects [83]. Although it is traditionally believed that the activated microglia and macrophages are difficult to distinguish from one another, recent nanotechnology can specifically regulate these inflammatory cells [25,26]. Jeong et al. investigated immune-modifying nanoparticles composed of carboxylated PLGA nanoparticles (500 μm) in a mouse SCI model and they demonstrated that carboxylated PLGA immune-modifying nanoparticles can limit infiltration of circulating inflammatory monocytes into the injury site without affecting the resident microglia [25]. In another study, Papa et al. investigated the effect of poly-ε-caprolactone-based nanoparticles loaded with minocycline (PCL-mino-NPs) on the inflammatory response associated to microglial cells in SCI. They demonstrated that the PCL-mino-NPs can be taken up to a large extent in microglia cells but minimally in recruited monocyte-derived macrophages [26].

Although several studies showed that transplantation of prestimulated macrophages into rodent SCI models improved functional recovery [84,85], a Phase II randomized controlled trial did not find a significant difference between an autologous macrophage therapy group and a control group [86]. This failed trial suggests that prestimulated macrophages might change their polarization after being transplanted in vivo so that direct modulation of the macrophage in vivo could be necessary. Several nanomaterials have been applied to deliver drugs into macrophages during in vivo studies [26–28]. The PCL-mino-NPs can efficiently decrease the activation of resident microglial cells, reduce the inflammatory response and improve the locomotor function up to 63 days post injury after acute injection following SCI compared with untreated SCI control mice [26]. Etanercept (ET), an inhibitor for TNF-α, was reported to be loaded in synthesized molybdenum disulfide PEG (MoS₂@PEG) nanoflowers to treat SCI. In vitro assay demonstrated that an ET-loaded vehicle obviously inhibited M1 markers (TNF-α, CD86 and iNOS), while promoting the expression of M2 markers (Agr1, CD206 and IL-10). In a mouse SCI model, ET-MoS₂@PEG effectively directed macrophages toward M2 polarization, reduced the extent of injured areas and significantly improved locomotor recovery [27]. In another study, poly(methyl methacrylate) nanoparticles were reported to enter specifically into activated microglia/macrophages and release compounds, as demonstrated by TO-PRO-3™ staining in vivo [28]. Being a fluorescent molecule for studying the nanocarrier properties, treatment with TO-PRO-3 does not induce difference from the SCI groups.

Other therapeutic agents can also be delivered in nanomaterials to regulate macrophage polarity. Because the transcription factor IRF5 can upregulate genes associated with M1 macrophages, researchers tried to silence the Irf5 gene in macrophages by intravenous injection of an siRNA-containing lipidoid nanoparticle. The results showed that silencing of Irf5 in a mouse SCI model resulted in both a dramatic M1 decrease and a significant M2 increase in the injury site, accompanied by a significant improvement of locomotor function [29].

Oxidative stress

Mitochondrial dysfunction has been shown to be crucial for the progression of secondary injury and neuronal cell death. In normal conditions, constitutively produced O₂ from mitochondrial metabolism is transformed by endogenous antioxidants such as superoxide dismutases. Physiologically, there is a dynamic balance between reactive oxygen and nitrogen species (ROS/RNS) generation and scavengers. Following traumatic injury, levels of ROS/RNS are rapidly elevated due to increased production beyond the normal capabilities of antioxidant systems and can further interact with lipids, nucleic acids and proteins to cause changes in the function of cellular components and, ultimately, cellular damage (Figure 4) [87–90].

Figure 4. . Oxidative stress in injured spinal cord after spinal cord injury.

NO: Nitric oxide; ROS: Reactive oxygen species.

Isaksson et al. evaluated the role of inducible nitric oxide synthase (iNOS) on the functional outcome following SCI in knockout mice lacking the iNOS gene (iNOS^−/−); they reported that these mice clearly tended to have a better functional outcome than wild-type mice at 14 days post injury [91]. Therefore many researchers have been focused on reducing oxidative stress, ROS/RNS and NO production following SCI by delivering therapeutics using biomaterial-based delivery systems. Luo et al. reported that PEG, a hydrophilic polymer, can restore membrane integrity of an injured spinal cord by resealing the injured membrane after compression injury. They also reported that PEG has been shown to significantly suppress injury-induced ROS elevation and lipid peroxidation levels due to the membrane repair mechanism [30].

In another study, Cho et al. also reported that PEG can rapidly restore membrane integrity, reduce oxidative stress, restore impaired axonal conductivity and mediate functional recovery in rats, guinea pigs and dogs. However, they encountered limitations related to the concentration and the molecular weight of PEG for the application. Therefore they developed PEG-decorated silica nanoparticles (PSiNPs) and evaluated their effect on membrane-sealing ability and neuroprotective effects after spinal cord trauma. They reported that PSiNPs significantly reduced the formation of ROS and the process of lipid peroxidation of the membrane. They also reported that recovered conduction through the cord lesion in the PSiNP-treated group was much better than in the control group [31]. Nukolova et al. tried to deliver SOD1, an efficient ROS scavenger, to the lesion site to mitigate SCI-induced oxidative stress and tissue damage. They developed double-coated nanozymes by coating with anionic block copolymer, PEG-polyglutamic acid after cross-linking the polyion complex of SOD1 to poly-cation PEG-polylysine (single-coat nanozyme). They reported that the double-coated nanozymes retained their enzymatic activity and ability to protect SOD1 under physiological conditions and that a single intravenous injection of double-coated nanozymes (5 kU of SOD1/kg) produced improved locomotor function in a rat moderate SCI model [32].

In other studies, Li et al. developed zonisamide-loaded triblock copolymer nanomicelles. The antiepileptic drug zonisamide (ZNS) was loaded in triblock monomethyl PEG-poly(l-lactide)-poly(trimethylene carbonate) copolymer and ZNS-loaded nanomicelles (ZNS-Ms) were intravenously injected in hemisection SCI rats. The released ZNS from the ZNS-Ms showed significant antioxidative and neuroprotective effects as well as producing improved motor functional recovery compared with animal groups injected with free ZNS or vehicle only [33]. Similarly, iron oxide nanoparticles were shown to scavenge the free radicals after T11 complete transection injury in rats. Histological analysis showed reduced lesion volume of spinal cord tissue and significant functional motor and sensory recovery [34].

In an alternative approach to reducing oxidative stress after SCI, some researchers have also conducted studies to remove acrolein using hydrazine. Acrolein is a highly reactive aldehyde produced after CNS injury and a very potent endogenous toxin. It can be inactivated by the formation of hydralazine adjuncts on the basis of interaction between acrolein and its trapping agent, hydralzine. Cho et al. developed mesoporous silica nanoparticles with encapsulated hydralazine and functionalized with PEG. They reported that these nanoparticles showed neuroprotection to cells damaged by a usually lethal exposure to acrolein by restoring their disrupted cell membrane and improving mitochondrial function associated with oxidative stress in an in vitro acrolein-mediated cell injury model using PC-12 cells [35]. Later, they also developed a hydrazine-loaded chitosan nanoparticle to potentially provide dual benefit by using chitosan as a membrane sealer (similar to PEG) as well as hydrazine to ameliorate the damaging effects of acrolein exposure. They reported that this hydrazine-loaded chitosan nanoparticle reduced damage to membrane integrity, secondary oxidative stress and lipid peroxidation in an in vitro acrolein-mediated cell injury model using PC-12 cells [36].

SCI animal models

Both small and large animal models are used to study SCI pathophysiology and investigate therapeutic strategies to improve functional recovery. Several reviews describing and comparing different SCI animal models have recently been published [92–95]. The most suitable choices for experimental SCI are rats and mice due to their size and cost, even though there are differences in anatomical, physiological and behavioral responses to injury between humans and rodents. One important difference between rat and mouse SCI models is that a modest degree of regeneration can occur at the lesion site in the mouse model that is not observed in the rat. Another important difference between the two is that a fluid-filled cystic cavity is formed after SCI in the rat model, which is consistent with human pathophysiology, while no cystic cavity formation occurs in the mouse model. For these reasons, an animal model of SCI should be chosen carefully based on the objectives of the study. Sharif-Alhoseini et al. published a systematic review of animal SCI models, analyzing study objectives, model species, injury site/method and outcome. The two most common study objectives were testing drug and growth factor interventions (36.6%) and investigating pathophysiological changes (30.2%). Rats were the most widely used species (72.4%) and the thoracic region the most common injury site (81%). Injury models included contusion (41%), transection (32.5%) and compression (19.4%) injuries, and the majority of studies (>50%) used both anatomical and behavioral assessment [92]. The authors concluded that the choice of animal model should be based on the study objectives and that contusion/compression models more closely resemble human injury, while transection models allow more unequivocal assessment of anatomical regeneration. Rodent models are cost effective and useful for discovery research and proof of principle, but there is a critical need for a transitional model between rodents and humans.

Large animals such as dogs, cats, pigs and nonhuman primates can be used as preclinical models for SCI. The first documented SCI canine model was described by Allen in 1911 [96] and this model has been used to study pathology [97] and functional outcomes [98,99] after SCI. Cats are commonly used to study locomotion following SCI [100–102]. Pigs have been a popular preclinical model due to their physiological similarity to human [103,104]. Nonhuman primates may be most useful as a preclinical SCI model due to their genetic, anatomical and physiological closeness to humans [105,106]. However, these large animal models have limitations due to size, cost, animal care and ethics, especially with respect to nonhuman primates. Nonetheless, animal models can be very important for understanding injury mechanisms, providing information on anatomical plasticity and pathophysiology of SCI and developing therapies for SCI; selection of the appropriate animal model can be the critical factor for a successful study.

Although encouraging results have been obtained in many preclinical animal studies, several barriers hinder their translation to successful human clinical trials. These include: differences in anatomy, physiology, pathology and function between experimental animal models and humans; limited standardization among SCI injury models and outcome measures; challenges with reproducibility of preclinical results; and the fact that most preclinical studies apply interventions during the acute phase of an SCI, while most human patients are in the chronic phase. Increasing success in progression from preclinical studies to clinical trials will require improved standardization and reproducibility and more widespread adoption of transitional models that bridge the pathophysiological gap between rodents and humans.

Conclusion

Continued damage to nerve cells after the primary injury can result in a secondary SCI. There are many therapeutic strategies to inhibit these secondary injuries, but one of the major hurdles is to deliver therapeutics in the injury site through the blood–brain and blood–spinal cord barriers. Here we summarized the transport of therapeutic targets using nanomaterials as a delivery system to mitigate secondary injury, such as BSCB disruption, inflammatory response and oxidative stress. Although this review may not cover all the nanomaterial-based therapeutic strategies for SCI, the research we have discussed shows the potential of nanomaterial-based delivery systems for more efficient treatment of the complex pathologies after SCI by delivering single or multiple therapeutics.

Future perspective

Secondary injury cascades after a primary insult are very complex processes and offer a therapeutic window for pharmacological treatments to prevent progressive tissue damage and improve functional outcomes. In this review, we discussed the potential of nanomaterial-based therapeutic strategies for the reduction of secondary injury in preclinical SCI animal models. We expect continued improvement in the functionality of nanomaterial-based delivery systems, improving combinatorial therapies by delivering multiple therapeutics to treat the complex pathologies of SCI.

Executive summary.

Background

• Spinal cord injury (SCI) is a major cause of morbidity and mortality throughout the world.

• Secondary injury after SCI occurs from hours to months after the primary insult and this delayed nature of secondary injury provides a therapeutic window.

• Nanomaterial-based delivery systems are a promising direction for the transport of therapeutic targets to mitigate secondary injury after SCI.

• This review discussed nanomaterial-based therapy, specifically targeting blood–spinal cord barrier (BSCB) disruption, mitigation of the inflammatory response and reduction of oxidative stress after SCI.

Hyperpermeability of the blood–spinal cord barrier

• SCI results in BSCB damage, which increases microvasculature permeability, allows extravasation of immune cells and increases the inflammatory response around the injury site.

• Several strategies using nanowire and solid lipid nanoparticle drug-delivery systems have been investigated for repairing the BSCB and improving neurological outcomes.

Inflammatory response

• Reducing neutrophil infiltration or inhibiting microglial activation can significantly improve the prognosis of SCI.

• The steroidal anti-inflammatory drug methylprednisolone was approved to treat acute SCI, but concerns over efficacy and side effects following high-dose systemic administration have reduced its use.

• Several nanocarrier strategies to locally deliver methylprednisolone and other anti-inflammatory drugs have been developed and were shown to reduce infiltration of immune cells and production of inflammatory cytokines.

Oxidative stress

• Mitochondrial dysfunction is crucial for the progression of secondary injury and neuronal cell death, and levels of reactive oxygen and nitrogen species are rapidly elevated after traumatic spinal cord injury.

• Nanoparticle delivery systems have been introduced to reduce oxidative stress and tissue damage.