Background
Although inflammation is known to be the major pathology of multiple sclerosis (MS), the mechanisms underlying tissue damage and functional loss remain unclear [
1,
2]. While myelin degeneration has long been considered the primary neuropathological characteristic for MS, recent studies indicate that axonal degeneration is also an important component of the pathology [
3‐
5]. In fact, there is strong evidence suggesting that MS is a neurodegenerative diseases [
3,
6‐
8]. Indeed, the integrities of both myelin and axons are indispensable for neuronal function and survival [
9]. Therefore, either myelin or axonal damage could theoretically lead to axonal conduction loss and degeneration seen in MS [
10‐
13]. Consistent with this notion, it has been suggested that axonal disruption may represent irreversible neurodegeneration in patients with MS [
3]. This may in part explain why conventional strategies focusing solely on myelin protection have resulted in few effective treatments to slow or prevent MS progression [
1].
Despite its potential importance in MS, axonal damage has attracted significantly less attention compared to myelin damage while both are known to lead to neurodegeneration in MS [
6,
7,
14]. Consequently, the pathological role of axonal damage in MS remains insufficiently characterized. Specifically, the key cellular processes that trigger axonal degeneration remain unclear. We have previously shown that axonal membrane damage contributes to axonal degeneration observed in CNS trauma [
11‐
13,
15‐
17]. We have also shown that acrolein, a pro-inflammatory aldehyde that is capable of inflicting axonal membrane damage and functional loss [
18‐
24], is elevated and likely plays an important pathological role in MS [
25]. In light of this evidence, we speculate that damage to the axonal membrane, or axolemma, likely leads to neuronal degeneration and loss of neurological function, and therefore contributes to the development and progression of symptoms observed in MS.
Polyethylene glycol (PEG), a hydrophilic polymer, is well known for its ability to seal neuronal membranes and consequently restore integrity and associated neuronal function [
13,
26‐
29]. In particular, it has been shown that PEG is capable of repairing axolemmal damage and provide neuroprotection in traumatic spinal cord injury [
19,
26‐
28,
30‐
40]. However, the therapeutic effect of PEG has not been examined in non-traumatic CNS illnesses, such as MS, in which axonal membrane damage likely plays a role leading to axonal degeneration. Therefore, the primary focus of this study was to determine whether axolemmal disruption can be detected and to examine its possible correlation with functional deficits associated with MS. Subsequently, we also aimed to confirm the pathological role of axolemmal disruption in MS and to assess the therapeutic efficacy of administering PEG as a membrane sealant.
Discussion
Based on the current study using the HRP-exclusion assay, a well-established method of assessing axonal membrane permeability [
11,
13,
15‐
17,
30,
42], we have determined that there is a significant level of axonal membrane disruption in EAE mice when symptoms peak. In addition, we also noted a prominent increase, although at a lesser degree, of axonal membrane damage in the pre-symptom stage in EAE mice (Fig.
2). To our knowledge, this is the first evidence of axonal membrane damage in EAE mice, an animal model MS.
Furthermore, PEG, a known membrane repairing agent [
28‐
31], when applied daily for 4 weeks after induction, can significantly reduce axonal membrane disruption in EAE mice (Fig.
2). The restoration of axonal membrane integrity by PEG was accompanied by significant amelioration of behavioral deficits, including a delay of the onset of motor deficits typical of EAE, a delay of emergence of peak symptom, and a reduction of peak symptom severity (Figs.
3 and
4). Taken together, we have presented initial evidence that axonal membrane disruption is an important feature of the overall pathology in EAE mice that is at least in part responsible for the behavioral deficits. Our study also presents the first indication that PEG could be used as a therapeutic agent to effectively repair axolemma, arrest axonal degeneration and reduce motor deficits in EAE mice.
Although not examined in this study, it is likely that PEG-mediated axonal membrane repair also leads to the reduction of axonal degeneration, a known pathology of MS [
3,
4,
43]. It is well known that axonal membrane disruption, if not repaired, will lead to axonal degeneration, neuronal cell death and overall neuronal tissue degeneration [
13,
17,
44‐
46]. It has also been demonstrated repeatedly that PEG-mediated neuronal membrane repair can lead to the reduction of oxidative stress and mitochondrial dysfunction which are known contributors to axonal degeneration and neuronal cell death [
19,
26,
29,
35,
47]. As such, PEG may also provide neuroprotection by indirectly suppressing oxidative stress and inflammation. Therefore, we postulate that PEG-mediated membrane repair can mitigate axonal degeneration and could promote a range of cellular functions that lead to the improvement of motor function in EAE mice.
Although we did not confirm the presence of PEG inside the spinal cord in the current study following systemic application, we believe the main location of PEG treatment is in the central nervous system (CNS), particularly in spinal cord. This is because we have detected significant membrane repair in spinal cord when PEG was applied comparing to no PEG (Fig.
2), and PEG is known to be able to reach spinal cord following systemic application [
48]. It is unlikely that PEG-mediated neuroprotection is due to peripheral effects, considering the main pathology of EAE is in CNS [
2].
Despite the strong evidence of severe axonal membrane damage in EAE, the mechanisms of such axolemmal damage remain to be elucidated. Based on the previous studies from our and other labs, we suggest that acrolein, a lipid peroxidation byproduct, may be one of the culprits. In a recent study from our lab it was shown that acrolein was increased significantly in EAE mice [
25]. We have also shown that acrolein can cause membrane disruption in various preparations at levels that are likely achievable in in vivo pathological conditions [
20‐
22,
49‐
51]. In fact, acrolein has been suggested to cause neuronal damage in trauma by disrupting neuronal membrane through a delayed mechanism [
24,
49,
50,
52‐
54]. In light of this evidence, we hypothesize that acrolein may play a role in axonal membrane disruption in EAE mice. One critical piece of evidence supports this hypothesis is that hydralazine, an effective acrolein scavenger, can lower acrolein levels and reduce motor deficits in EAE mice [
25]. In addition, a recent study from our group demonstrated that acrolein-mediated axonal conduction loss can be partially mitigated by a potassium channel blocker, indicating a concomitant acrolein-mediated myelin damage in addition to axonal lesions [
24]. This is because augmented potassium channel activity is a known consequence of myelin damage in injured axons [
55]. Consistent with such notion, we also have found that acrolein trapping treatment was associated with restoration of neuronal membrane integrity, reduced neurodegeneration and enhanced functional recovery in traumatic spinal cord injury [
21,
22,
24,
50,
51,
53]. It will be interesting to confirm the likely scenario that anti-acrolein therapy alone could lead to the preservation or restoration of axonal membrane integrity in EAE.
In the current study, in addition to the severe membrane disruption observed when symptoms peak, we also noted a less severe, yet still significant level of membrane disruption, and therefore neurodegeneration, in the pre-symptom period defined as a week prior to the emergence of motor deficits (Figs.
1 and
2). Therefore, significant level of membrane damage and neurodegeneration appear to already exist in the pre-symptom period while no concomitant noticeable behavioral deficits were detected. This phenomenon may be explained by the fact that there is a significant amount of implicit redundancy of axons to support neuronal function. Therefore, there is likely a threshold level of axonal damage and neurodegeneration that must be reached before observing behavioral changes. This would justify the notion that membrane damage could start before the onset of behavioral deficits and that the emergence of behavioral deficits signifies a critical level of axonal damage and degeneration. Hence, the initial membrane damage could theoretically serve as an indication to predict the onset of behavioral deficits at a later date.
In light of these observations related to the relationship between axonal membrane damage and behavioral performance, we suggest that axonal membrane damage in MS could serve as an important diagnostic measurement. First, it can potentially be used as a biomarker for the detection of MS before the emergence of symptoms. This could help to institute earlier treatments to deter neurodegeneration for better therapeutic outcomes provided that the pre-symptomatic axonal membrane damage can be firmly identified. Second, axonal membrane integrity could also serve as an indication of the effectiveness of membrane repair therapy such as PEG. Such a strategy is understandably dependent on development of a reliable non-invasive imaging method to detect axonal membrane disruption or degeneration.
It has become increasingly clear that, in addition to myelin damage, axonal degeneration may also play a critical role in EAE pathology. From a basic cellular biology view, the integrity of both myelin and the axonal membrane are essential for axonal conduction. Therefore, axonal degeneration along with myelin damage, are likely to be equally important contributing factors in axonal conduction failure and behavioral deficits in EAE mice [
5,
12,
15,
17,
54,
56,
57]. The current study further highlighted the critical role of axonal damage, and axolemmal disruption in particular, in neurodegeneration and the functional loss in MS. To further stress the importance of axonal degeneration in MS, there is evidence that axonal damage and neurodegeneration may be the main cause of functional loss without obvious myelin damage in some human case of MS. For example, based on a histopathologic investigation using autopsy tissue from MS patients, Trapp and his colleagues have shown compelling evidence of axonal degeneration in the absence of myelin loss [
43]. Therefore, axonal damage, a key feature of neurodegeneration, may not just be one of the important compounding pathological factors, but rather it may be among the primary and critical factors that are sufficient to cause clinical functional deficits in MS. To further highlight the importance and causal role of axonal damage in MS pathology, we have noted a significantly higher susceptibility of acrolein-mediated axonal damage compared to acrolein-mediated myelin damage. Specifically, Shi and his colleagues have found that using an ex vivo preparation of extracted rodent spinal cord segment, the threshold of acrolein needed to inflict membrane damage is two magnitudes lower than that needed to cause significant myelin damage [
10,
49]. This suggests that in MS patients, axonal degeneration may be, in some cases, the primary pathology that precedes myelin damage. Furthermore, due to the existing evidence of damaging both axon and myelin, acrolein may be a critical link for the damage of both myelin and axons, two known pathologies in EAE. This hypothesis is supported by the factor that acrolein scavenging could mitigate the damage of both myelin and axons [
25,
26,
29].
The emergence of the importance of axonal damage as the critical pathology warrants closer examination of our existing therapeutic strategies as well as our efforts to establish new therapies. It appears reasonable to suggest that a treatment regimen should include axonal repair and protection in addition to myelin protection. This combination of treatments may be a synergistic treatment strategy and could result in increased efficacy. We have previously shown that both anti-acrolein (by hydralazine) and membrane repair (by PEG), when used alone, can offer significant, albeit partial alleviation of behavioral deficits [
25] (Fig.
3). Furthermore, though PEG significantly delayed the onset and reduced severity in most of the symptomatic periods, such symptom reduction was temporary. PEG treatment did not lead to significant behavioral improvement beyond 26 days post induction, despite effective sealing of the axonal membrane. Therefore, it appears that PEG-mediated membrane repair alone cannot offer long-term symptom reduction. This could be the case for at least two reasons. First, this may indicate that membrane repair strategies need to be combined with approaches that remove causes of cellular damage, such as acrolein, to ultimately protect the cell. Second, although a proven axonal membrane repair agent, PEG has not been shown to also repair myelin damage.