Remyelination strategies in multiple sclerosis: a critical reflection

Abstract

Remyelination is the natural repair mechanism of demyelination and can be a highly efficient process in multiple sclerosis. However, in the majority of lesions, this regenerative approach is incomplete or fails. It is believed that remyelination protects against progressive axonal damage and thus long-term disability in patients with multiple sclerosis. For this reason, therapeutic promotion of remyelination represents an attractive option for preventing disease progression. In this editorial we casts a critical eye over the most frequently used experimental settings which aim to uncover potential remyelination promoting drugs. This article reflects upon the personal opinion of the author who currently used animal models allow to assess the potency of pharmacological interventions to accelerate, but not to induce myelin repair. Furthermore, it is discussed how remyelination and neuroprotection might well be two separate entities. Thus, induction of remyelination does not necessarily prevent disease progression in multiple sclerosis patients.

KEYWORDS:

• remyelination
• multiple sclerosis
• neuroprotection
• axonal damage
• oligodendrocytes
• anima models

Demyelination is still a characteristic entity for the pathology of multiple sclerosis (MS) but it has become increasingly apparent in recent years that substantial axonal and neuronal loss are equally important features. Remyelination is one of the best-documented and most robust examples of tissue repair in the human central nervous system (CNS). Whereas in adult CNS, the regeneration of destroyed neurons and their axonal processes are very limited, lost myelin sheaths can be very effectively repaired. Approximately 20–30% of postmortem MS tissues demonstrate remyelination which can occur early and late during the course of the disease.[1,2] Steps involved in this regenerative approach include activation and proliferation of oligodendrocyte progenitor cells (OPC)s, migration of these OPCs towards the demyelinated axons and OPC interaction with the axon which culminates in OPC differentiation and remyelination. Of note, the source of such OPCs might be manifold including progenitor cells dispersed in the brain parenchyma or located within neurogenic niches.[3]

The beneficial effects of remyelination are well-known and include the restoration of axonal conduction properties that are lost following demyelination [4] as well as axonal protection.[5] The later beneficial aspect of remyelination is believed to be at least in part because of trophic axonal support from glia cells.[6] Additionally, neurons benefit from remyelination because in electrically active but demyelinated fibers, the restoration of ion gradients by the Na⁺, K⁺-ATPase consumes a large fraction of available ATP. Remyelination strongly reduces this energy consumption and thus can be neuroprotective. Two important aspects of remyelination will be addressed in this editorial: first, do we use appropriate animal models to develop new remyelination inducing drug targets and second, is remyelination and neuroprotection indeed intimately linked.

The majority of available immunomodulatory medications for MS are approved for the relapsing-remitting (RR) course of the disease, for which they reduce relapse rate, MRI measures of inflammation and the accumulation of disability. Approved medications include β-interferons (Avonex, Betaseron, Extavia and Rebif), fingolimod (Gilenya), glatiramer acetate (Copaxone), mitoxantrone (Novantrone), natalizumab (Tysabri), alemtuzumab (Lemtrada), teriflunomide (Aubagio) and dimethyl fumarate (Tecfidera). These medications are, however, of little benefit during progressive MS where axonal degeneration following demyelination outweighs inflammation. For obvious reasons, this discrepancy in therapeutic efficacy of currently approved drugs has sparked great interest in the development of new remyelination therapies aimed at providing neuroprotection and functional recovery. The most commonly used animal models used to study MS related aspects of remyelination are toxin models. In principal, toxin-mediated demyelination with subsequent remyelination can be induced by either focal injection or systemic administration of the toxin. While focal demyelination is usually induced by injection of lysolecithin (also called lysophosphatidylcholine; LPC) or ethidium bromide,[7] the copper-chelator cuprizone is used for systemic administration.[8] Although it is beyond the scope of this article to describe the specific features of both models in detail, it is worth to mention that in both models robust endogenous remyelination after toxin-induced demyelination takes place, at least if “conventional” protocols are used. In the cuprizone model, complete and fast remyelination follows after a 5-weeks intoxication period (i.e., called acute demyelination).[9] After focal, LPC-induced demyelination, endogenous remyelination is as well robust and occurs within 2–3 weeks.[10,11] It is important to notice that under such experimental paradigms, we cannot investigate the potency of a pharmaceutical compound to induce remyelination, but rather can assess the potency of drugs to accelerate remyelination. It is clear from extensive pathological studies that remyelination is insufficient in a majority (around 80%) of MS lesions.[2] Although it is principally possible that remyelination fails because the repair process per se is too slow and accelerating this process might well be beneficial for neuronal health, I consider it more likely that the environment is not supportive for myelin repair which could, among other mechanisms, impair OPC differentiation.[12] Of note, a non-supportive environment includes an inflammatory state acting directly on the wealth of OPCs, their differentiation and migration.[13] As a conclusion, in the field of remyelination therapy, the translation from bench to bedside might be hampered because of the discrepancy between the reason of remyelination failure in MS patients and the high endogenous remyelination capacity in the applied animal models. In this context it is important to recognize that there are animal models/experimental setups available which indeed show impaired endogenous remyelination capacity. First, despite of acute demyelination, chronic demyelination can be induced in the cuprizone model which provides, in contrast to the acute protocol, an environment with impaired endogenous remyelination .[5,14,15] Furthermore, the process of remyelination is as well hindered in aged animals in both, the cuprizone and LPC-model.[16–18] Using these paradigms would allow us to study the induction rather than the acceleration of remyelination.

The second remyelination dogma is that remyelination is neuroprotective and several of the suggested underlying mechanisms have been listed above. Some important new aspects of the relationship between remyelination and acute axonal damage have recently been established using the cuprizone model. Following cessation of cuprizone treatment (5 weeks), animals showed an initial recovery of locomotor performance. However, long after remyelination was completed (approximately 6 months after the last demyelinating episode), locomotor performance again declined in remyelinated animals compared with age-matched controls.[19] Ongoing axonal damage during the phase of remyelination was well reported by other groups.[5,14] These studies clearly highlight that remyelination and axonal degeneration might occur independently from each other. Obviously, remyelination alone may not fully compensate for neuronal stress which occurs from a period of demyelination. To conclude, even after completed remyelination, axonal degeneration can continue to progress at a low level, accumulating over time, and that once a threshold is passed axonal degeneration can become functionally apparent in the long-term. It is therefore wise not just to focus on remyelination as an aspect of neuroprotection in MS but as well follow direct neuroprotective strategies. In either case, adequate animal models have to be applied.

One substantial benefit of the cuprizone and focal LPC-model with regard to remyelination studies is that the spatial and temporal distribution of demyelinated lesions is highly predictable. This does not apply for the various EAE models, where foci can be virtually everywhere within vulnerable regions such as within the spinal cord or the cerebellum. On the other hand, in the cuprizone as well as in the LPC-model, lesions develop with little, if any, autoimmune reaction. Thus, while both models allow us to study straight forward the principles of OPC activation, migration and differentiation, the impact of inflammation on these events cannot adequately be addressed. Other animal models, such as chronic EAE or Theiler’s murine encephalomyelitis virus (TMEV)-induced demyelination present with limited remyelination and significant immune system involvement that may better represent the complex pathology of MS. Of note, unpredictable lesion distribution is as well a matter in these models.

Several candidate drugs with the potential to promote remyelination in MS patients have been identified in the past couple of years,[11] and new in vitro screening approaches have been developed.[20] Although far from clinical use, it is now time to test these candidate drugs in appropriate remyelination animal models before the subsequent initiation of clinical trials. Although acceleration of remyelination might be important, induction of remyelination in the non-permissive environment should be in the focus. Furthermore, since remyelination and neuroprotection might be two separate nuts to crack, neuroprotective molecules will likely be combined in the future with efforts to promote remyelination. A combination of the three components immunomodulation, neuroprotection and remyelination is as well a promising strategy.

First steps in the right direction have been made with, for example, anti-LINGO-1 antibodies. A Phase II trial in people with acute optic neuritis (named RENEW) demonstrated an improvement in the study’s primary endpoint, that is, recovery of optic nerve latency (time for a signal to travel from the retina to the visual cortex) relative to placebo. However, secondary endpoints, such as a change in thickness of the retinal layers or visual function were not met in this clinical trial. Results of the SYNERGY study, which combines anti-LINGO-1 therapy with Avonex^® are anticipated in the year 2016.

Other clinical trials are ongoing and will show the potency of pro-remyelinating strategies to provide long-term benefit for MS patients.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Notes on contributors

Markus Kipp