Remyelination in the Central Nervous System

Abstract

The inability of the mammalian central nervous system (CNS) to undergo spontaneous regeneration has long been regarded as a central tenet of neurobiology. However, while this is largely true of the neuronal elements of the adult mammalian CNS, save for discrete populations of granule neurons, the same is not true of its glial elements. In particular, the loss of oligodendrocytes, which results in demyelination, triggers a spontaneous and often highly efficient regenerative response, remyelination, in which new oligodendrocytes are generated and myelin sheaths are restored to denuded axons. Yet remyelination in humans is not without limitation, and a variety of demyelinating conditions are associated with sustained and disabling myelin loss. In this work, we will (1) review the biology of remyelination, including the cells and signals involved; (2) describe when remyelination occurs and when and why it fails, including the consequences of its failure; and (3) discuss approaches for therapeutically enhancing remyelination in demyelinating diseases of both children and adults, both by stimulating endogenous oligodendrocyte progenitor cells and by transplanting these cells into demyelinated brain.

IDENTIFYING REMYELINATION IN ANIMAL MODELS

Remyelination is the process in which new myelin sheaths are restored to axons that have lost their myelin sheaths as a result of primary demyelination (Franklin and ffrench-Constant 2017).⁶ Primary demyelination is the term used to describe the loss of myelin from an otherwise intact axon and should be distinguished from myelin loss secondary to axonal loss—a process called Wallerian degeneration or, misleadingly, secondary demyelination. Remyelination is sometimes referred to as myelin repair. However, this term suggests a damaged but otherwise intact myelin internode being “patched up,” a process for which there is no evidence, and which does not emphasize the truly regenerative nature of remyelination, in which the pre-lesion cytoarchitecture is all but fully restored. Remyelinated tissue very closely resembles normally myelinated tissue but differs in one important aspect—the newly generated myelin sheaths are typically shorter and thinner than the original myelin sheaths. When myelin is initially formed in the peri- and postnatal periods, there is a striking correlation between axon diameter and myelin sheath thickness and length established during myelination, which is less apparent in remyelination. Instead, myelin sheath thickness and length show little change with increasing axonal diameter with the result that the myelin is generally thinner and shorter than would be expected for a given diameter of axon (Fig. 1). Although some remodeling of the new myelin internode occurs, the original dimensions are rarely regained (Powers et al. 2013). The relationship between axon diameter and myelin sheath is expressed as the G ratio, which is the fraction of the axonal circumference to the axon plus myelin sheath circumference. The identification of abnormally thin myelin sheaths (> than normal G ratio) remains the “gold standard” for unequivocally identifying remyelination and is most reliably identified in resin-embedded tissue, viewed by light microscopy following toluidine blue staining, or by electron microscopy. This effect is obvious when large diameter axons are remyelinated but is less clear with smaller diameter axons such as those of the corpus callosum, where G ratios of remyelinated axons can be difficult to distinguish from those of normally myelinated axons (Stidworthy et al. 2003). How is the relationship between myelin parameters and axon size established in myelination, and why is it disengaged in remyelination? In the peripheral nervous system (PNS), axonally expressed neuregulin (NRG)1-type III plays a key role; reduced expression results in thinner myelin sheath (increased G ratio), whereas overexpression leads to a thicker than expected myelin sheath (decreased G ratio) (Michailov et al. 2004). In the central nervous system (CNS), however, the role of neuregulins in controlling myelin sheath length and thickness is less clear (Brinkmann et al. 2008), although they play a role of rendering axons dependent on electrical activity for myelination. The factors that govern the G ratio in remyelination would seem to be distinct from those operating in developmental myelination, and an explanation for the increased G ratio in remyelination remain elusive. For example, overexpression of Nrg leads to CNS hypermyelination in development but not during remyelination (Brinkmann et al. 2008). Similarly, activation of the Akt pathway in the CNS, which results in thicker than expected myelin sheaths in development (Flores et al. 2008), does not result in thicker remyelinated sheaths following demyelination in the adult (Harrington et al. 2010). One hypothesis is that whereas the myelinating oligodendrocyte associates with a dynamically changing axon yet to achieve its full length and diameter, the remyelinating oligodendrocyte engages an axon that is comparatively static having already reached it mature size (Franklin and Hinks 1999). As a result, the remyelinating oligodendrocyte is not subjected to the same dynamic stresses encountered by the myelinating oligodendrocyte.

Figure 1.

The relationship between myelin sheath and the axon diameter in myelination and remyelination. (A) The relationship between the thickness of the myelin sheath (determined by the number of wraps or lamellae) and the axon diameter is expressed as the G ratio, calculated by dividing the diameter of axon by the diameter of the axon plus the myelin sheath. The higher the G ratio, the thinner the myelin sheath, where a G ratio of 1 corresponds to an un- or demyelinated axon. (B) In developmental myelination, there is an increase in myelin sheath thickness with increasing axonal diameter. In remyelination, however, the myelin sheath thickness remains the same regardless of the diameter. (This figure is based on data in Stidworthy et al. 2003 and Franklin et al. 2012; adapted and reprinted with permission from Franklin and ffrench-Constant 2017.)

IDENTIFYING REMYELINATION IN HUMANS

Identifying remyelination in humans has proven to be particularly challenging, as several of the imaging techniques proposed to measure myelin content changes in the CNS do not fully meet the required criteria of being sensitive and specific to myelin while also being reproducible and clinically meaningful. The key objective of myelin imaging applied to humans with demyelinating diseases is to quantitatively measure myelin loss and myelin regeneration, as well as to distinguish between repaired and partially demyelinated tissues.

Measuring Remyelination in Humans with Magnetic Resonance Imaging

Several advanced magnetic resonance imaging (MRI) techniques have been introduced to explore myelin content changes in humans, including magnetization transfer imaging (MTI), T2 relaxometry, and diffusion tensor imaging (DTI). Compared with histology, either in animal or in human studies, these techniques have been shown to present a reasonably accurate and comparable correlation with myelin content (Mancini et al. 2020). A few have been already employed to specifically measure remyelination in patients with multiple sclerosis (MS) (Bodini et al. 2021).

Changes in magnetization transfer ratio (MTR), an MTI-based metric reflecting the exchange rate between the bound protons to macromolecules and the unbound protons in free water, are thought to be highly sensitive to myelin content variations due to the significant contribution of myelin to the macromolecules involved in the MT phenomenon (Schmierer et al. 2004; Moll et al. 2011; Moccia et al. 2020). Significantly reduced MTR values in newly forming lesions, possibly reflecting acute demyelination, have been found in the white matter of patients with MS, followed by a progressive, although partial recovery over the following months, likely to indicate a subsequent phase of remyelination (Chen et al. 2008). More recently, MTR has been used to generate MS-patient-specific maps of demyelination and remyelination at the cerebral cortex (Derakhshan et al. 2014; Lazzarotto et al. 2022). A promising development of MTI is inhomogeneous MTI, a refined technique that selectively images tissues with long dipolar relaxation time components, such as myelin-rich structures, allowing a more specific measure of myelin content in healthy individuals and of myelin content changes in patients with demyelinating disorders (Duhamel et al. 2019; Lee et al. 2022). Another established technique with good sensitivity and specificity for myelin is myelin water imaging, which measures the myelin water fraction (MWF), defined as the quickly decaying signal arising from water trapped between myelin sheaths (Mackay et al. 1994; MacKay and Laule 2016). MWF has been employed to investigate myelin damage and repair in MS, as well as other diseases in which myelin pathology is suspected (MacKay and Laule 2016).

Besides MTI and myelin water imaging, myelin loss and regeneration have also been investigated in MS using changes in radial diffusivity (RD), a DTI-based metric describing microscopic water movements perpendicular to axonal tracts that had previously been shown to mainly reflect myelin content in experimental models of white matter damage (Song et al. 2002, 2005). In particular, changes in RD values in demyelinating lesions, possibly reflecting myelin loss and repair, have been found to be able to discriminate the functional outcome following optic neuritis and spinal cord relapses in patients with MS (Freund et al. 2010; Naismith et al. 2010).

Measuring Remyelination in Humans with Positron Emission Tomography

While being very sensitive to microstructural injury, all these MRI-derived metrics only allow an indirect and insufficiently specific measure of myelin content changes in humans, since they are significantly affected by other pathological processes such as inflammation, edema, and axonal loss (Petiet et al. 2019; Moccia et al. 2020). An alternative quantitative technique offering a more direct and specific access to the myelin compartment in humans is positron-emission tomography (PET) with myelin-binding radiotracers (Bodini et al. 2021). While several PET tracers, mainly belonging to the stilbene- and benzothiazole-derivative classes, have been shown to bind selectively to the CNS myelin structure in experimental models, so far only ¹¹C-PiB, ¹⁸F-florbetaben, ¹⁸F-florbetapir, and ¹¹C-MEDAS have been employed in clinical studies to explore myelin content changes in patients with MS (Bodini et al. 2021). Although a group-level analysis of all myelin PET studies in patients with MS indicated a decrease in tracer binding in white matter lesions compared to healthy controls, indicating demyelination, the use of individual lesion- or voxel-based analysis strategies revealed significant heterogeneity in the extent of myelin loss within the lesions (Bodini et al. 2016; Carotenuto et al. 2020; Pytel et al. 2020; Zhang et al. 2021; van der Weijden et al. 2022). The measure of dynamic myelin content changes over time with myelin PET revealed highly heterogeneous remyelination profiles across patients and lesions, which have been found to significantly correlate with neurological disability and brain atrophy (Bodini et al. 2016; Tonietto et al. 2022).

Taken together, MRI and PET studies exploring changes in myelin content in MS unequivocally indicate that myelin loss and repair are heterogeneous processes across patients. Moreover, demonstrating in vivo that individual profiles of remyelination are key determinants of atrophy and clinical evolution, these studies support the notion that an effective reparative response to demyelination can protect neurons from degeneration and, ultimately, patients with MS from experiencing an irreversible accumulation of clinical disability.

REMYELINATION IS THE NORMAL RESPONSE TO DEMYELINATION

Remyelination as a regenerative process shares many common features with regenerative processes occurring in other tissues of the body, and is the expected or default response to demyelination. The evidence for this comes from both experimentally induced and clinical demyelination. When demyelination is induced by toxins injurious to oligodendrocytes and myelin (for example, by dietary cuprizone or direct delivery of lysolecithin or ethidium bromide), then remyelination in white matter usually proceeds to completion, albeit in an age-dependent manner (Sim et al. 2002b). An interesting situation arises in the cerebral cortex where lengths of axon can remain unmyelinated during developmental myelination (Gibson et al. 2014). Using time-lapse, two-photon microscopy it has been shown that following cuprizone-induced demyelination of the adult mouse somatosensory cortex, the pattern of remyelination that occurs is distinct from the pre-demyelination pattern (Orthmann-Murphy et al. 2020). There is evidence that axons undergoing primary demyelination in experimental or clinical traumatic injury undergo complete remyelination, and that the persistence of chronically demyelinated axons is unusual (Lasiene et al. 2008). An exception is when demyelination is induced by or associated with the adaptive immune response, such as occurs in the autoimmune-mediated condition MS and in a laboratory animal model, experimental autoimmune encephalomyelitis (EAE). In this context, remyelination occurs in an environment intrinsically hostile to the oligodendrocyte lineage, although recent evidence has challenged the dogma that such an environment will be inimical to remyelination (Yilmaz et al. 2023). Thus, remyelination failure in MS (and EAE) is not inevitable, but rather a feature of the specific disease environment of MS (Mezydlo et al. 2023). Nevertheless, even in MS—a disease prototypically associated with failed or inadequate remyelination—remyelination can proceed to completion, and can be anatomically extensive (Patrikios et al. 2006; Patani et al. 2007; Goldschmidt et al. 2009; Piaton et al. 2009). Remyelination can also be extensive in EAE, and models with significant persistent demyelination are unusual (Linington et al. 1992; Hampton et al. 2008). Remyelination is especially efficient following demyelination of cerebral cortical gray matter, in both experimental models (Merkler et al. 2006) and clinical disease (Albert et al. 2007; Strijbis et al. 2017), although the reason for this is unclear.

REMYELINATION RESTORES FUNCTION AND PROTECTS AXONS

Remyelination restores saltatory conduction and reverses functional deficits (Smith et al. 1979; Jeffery et al. 1999; Liebetanz and Merkler 2006; Traka et al. 2010). Compelling evidence in support of functional restoration by remyelination is provided by an unusual demyelinating condition in cats, in which the reversal of clinical signs is associated with spontaneous remyelination (Duncan et al. 2009).

An additional and key function of remyelination is the protective effect it has on the underlying axon (Irvine and Blakemore 2008). Axonal and eventually even neuronal loss is a major cause of the progressive nature of chronic demyelinating disease, such as occurs in MS (Trapp and Nave 2008), and is likely primarily due to the absence of the myelin sheath, rather than the direct damage by inflammation that accounts for axonal loss in acute lesions. Thus, patients on appropriate immunosuppressive therapy and with apparently quiescent disease still show monotonically increasing disability and clinical progression, as these patients manifest persistent demyelination regardless of their lack of active disease. Indeed, remyelination is not the principal reason for the resolution of clinical signs following an acute relapse, which rather likely results from the resolution of inflammation, paired with adaptive responses by affected axons that serve to restore conduction.

Evidence that myelin is required for axon survival was first obtained using genetic mouse models, and is supported by subsequent studies of human pathology (Nave and Trapp 2008). Transgenic mice lacking the enzyme 2′,3′-cyclic nucleotide 3′ phosphodiesterase (CNP) or the myelin transmembrane protein proteolipid protein (PLP) show long-term axonal degeneration, even in the presence of myelin sheaths that are either ultrastructurally normal or show only minor abnormalities (Griffiths et al. 1998; Lappe-Siefke et al. 2003). Further analysis of the Plp mutant mice has revealed a disturbance in axoplasmic transport in the absence of PLP (Edgar et al. 2004) and has led to the identification of myelin-associated Sirtuin 2 as a potential mediator of long-term axonal stability (Werner et al. 2007). Myelin is also important for axon survival in humans, as patients with Pelizaeus–Merzbacher disease (PMD) caused by mutations in PLP show axon loss (Garbern et al. 2002), and studies of MS autopsy tissue show that axon preservation is seen in those areas where remyelination has occurred (although whether this is because remyelination has occurred where axons persist or that axons persist because of remyelination is unclear) (Kornek et al. 2000). Axon degeneration also occurs as a consequence of genetically induced, oligodendrocyte-specific ablation, even in Rag 1-deficient mice that have no functional lymphocytes (Pohl et al. 2011). These observations offer compelling evidence that axonal survival is dependent on intact oligodendrocytes, and that axonal degeneration in chronically demyelinated lesions can occur independently of inflammation. The nature of the “trophic” exchange from oligodendrocyte to axon remains to be fully elucidated but at least in part rests on transfer of energy metabolites from oligodendroglia to axons through monocarboxylate transporter 1 (MCT1) (Lee et al. 2012; Morrison et al. 2013; Schäffner et al. 2023), a dependency that increases with adult aging (Philips et al. 2021).

MECHANISMS OF REMYELINATION

OPCs Are the Principal Source of New Myelin-Forming Oligodendrocytes

In most instances following experimental demyelination in animal models, remyelination involves the generation of new mature oligodendrocytes. In the vast majority of cases, the new oligodendrocytes that mediate remyelination are derived from a population of adult CNS progenitor cells, most often referred to as adult oligodendrocyte progenitor cells (OPCs), that in humans arise largely from fetal forebears in the outer subventricular zone (SVZ) (Huang et al. 2020). More broadly, these cells are also referred to as glial progenitor cells (GPCs), since in humans they are bipotential for oligodendrocytes and astrocytes alike and retain neurogenic competence as well under appropriate conditions (Nunes et al. 2003). These multiprocessed, proliferating OPCs are widespread throughout the adult CNS, occurring in both the white matter and gray matter (3%–5% of cells in the adult human forebrain, and 5%–8% in rodents) (Horner et al. 2000; Dawson et al. 2003; Richardson et al. 2011). Adult and fetal OPCs share many similarities, although the adult cells have a longer basal cell cycle time and migrate less rapidly than do their fetal counterparts (Wolswijk and Noble 1989) in rodents as well as humans (Windrem et al. 2004). However, following inflammatory demyelination, OPCs become activated by the inflamed environment and assume a transcriptomic profile more akin to that of the replicating OPCs of the neonate, rather than that of the resting OPCs of the intact adult CNS (Moyon et al. 2015).

Evidence obtained using Cre-lox fate mapping in transgenic mice following experimental demyelination has shown that OPCs produce the vast majority of remyelinating oligodendrocytes (Tripathi et al. 2010; Zawadzka et al. 2010). Remyelinating oligodendrocytes may also come from the stem and progenitor cells of the adult SVZ, either from the progenitor cells contributing to the rostral migratory stream (RMS) (Nait-Oumesmar et al. 1999) or from the GFAP-expressing neural stem cells (NSCs) of the SVZ per se (Menn et al. 2006). However, the contribution that SVZ-derived cells make relative to that from local OPCs may be small, and their contribution to repair beyond the periventricular white matter is likely negligible (Kazanis et al. 2017).

Although experimental evidence indicates that oligodendrocytes themselves do not give rise to new remyelinating oligodendrocytes (Crawford et al. 2016), evidence from experimental models indicates that in some instances new myelin sheaths can be generated from oligodendrocytes that survive within areas of demyelination, albeit shorn of their original myelin sheath–associated processes (Duncan et al. 2018). This type of remyelination has been shown in cat, mouse, and zebrafish models of demyelination, and in the latter is rare compared to OPC-mediated remyelination and is often aberrant, involving, for example, myelin of neuronal cell bodies (Duncan et al. 2018; Bacmeister et al. 2020; Neely et al. 2022). Carbon dating studies have been used to suggest that this form of remyelination occurs in humans in the MS brain (Yeung et al. 2019). However, difficulties in unequivocally establishing whether demyelination and subsequent remyelination has occurred in the human brain make it difficult to be certain how extensive this form of remyelination really is in MS (Neumann et al. 2020). The foci within MS tissue with above background levels of de novo oligodendrogenesis suggest that OPC-driven remyelination is also occurring (Yeung et al. 2019). Remyelination by surviving oligodendrocytes nevertheless represents an interesting new form of remyelination deserving of further investigation (Franklin et al. 2021).

Remyelination Requires the Activation, Recruitment, and Differentiation of Adult OPCs

In response to injury, local OPCs undergo a switch from an essentially quiescent state to a regenerative phenotype. This activation is the first step in the remyelination process and involves not only changes in morphology but also up-regulation of several genes, many associated with the generation of oligodendrocytes during development such as the transcription factors Olig2, Nkx2.2, MyT1, and Sox2 (Fancy et al. 2004; Watanabe et al. 2004; Shen et al. 2008). The activation of OPCs is likely to be in response to acute injury-induced changes in microglia and astrocytes, two cell types exquisitely sensitive to disturbance in tissue homeostasis (Glezer et al. 2006; Rhodes et al. 2006). These two cell types, themselves activated by injury, are the major source of factors that induce the rapid proliferative response of OPCs to demyelinating injury (Fig. 2). This response is modulated by levels of the cell cycle regulatory proteins p27Kip-1 and Cdk2 (Crockett et al. 2005; Caillava et al. 2011) and is promoted by the growth factors PDGF and FGF (Woodruff et al. 2004; Murtie et al. 2005), Endothelin-1 (Gadea et al. 2009), and many other factors associated with acute inflammatory lesions and demonstrated to have OPC mitogenic activity in tissue culture (Vela et al. 2002). Semaphorins are important regulators of OPC migration following demyelination: semaphorin 3A impairs OPC recruitment to the demyelinated area, while semaphorin 3F overexpression accelerates not only OPC recruitment, but also remyelination rate (Piaton et al. 2011). The population of areas of demyelination by OPCs is referred to as the “recruitment phase” of remyelination and involves OPC migration in addition to the ongoing proliferation.

Figure 2.

The stage of oligodendrocyte progenitor cell (OPC)-mediated remyelination. Endogenous OPCs (adult progenitors) become activated in response to the change in environment triggered by demyelination. In the activated state, OPCs divide and migrate to populate the area of demyelination at a density much greater than the surrounding tissue during the recruitment phase. In a timely manner, recruited OPCs exit the cell cycle and differentiate into myelin-forming oligodendrocytes—the differentiation phase. Note, a number of OPCs remain undifferentiated to restore the pre-lesion density of OPCs.

For remyelination to be complete, the recruited OPC must next differentiate into remyelinating oligodendrocytes—the differentiation phase (Fig. 2). This phase encompasses three distinct steps—establishing contact with the axon to be remyelinated, expression of myelin genes and generation of myelin membrane, and, finally, wrapping and compaction to form the sheath. Despite these being fundamental properties of oligodendrocytes, we still have an incomplete understanding about how axo-glial contact is established and how this interaction then regulates, within each individual cell process, the morphological changes that constitute myelination. Nevertheless, some molecules have been shown to contribute to the regulation of differentiation, and it is clear that the differentiation of OPCs into myelinating oligodendrocytes in development and during the regenerative process share many similarities (Fancy et al. 2011). FGF plays a key role in inhibiting differentiation as well as promoting recruitment and thereby regulates the correct transition from the recruitment to the differentiation phases (Armstrong et al. 2002), and IGF-I is another factor that plays major roles in both processes (Mason et al. 2003). Semaphorin 3A, in addition to its role in OPC recruitment (Piaton et al. 2011), is also an inhibitor of OPC differentiation (Syed et al. 2011). LINGO-1, a component of the trimolecular Nogo receptor, has been found to be a negative regulator of oligodendrocyte differentiation in development (Mi et al. 2005), while mice deficient in LINGO-1 or treated with an antibody antagonist against LINGO-1 exhibited increased remyelination and functional recovery from EAE (Mi et al. 2007). The canonical Wnt pathway is as a very powerful negative regulator of oligodendrocyte differentiation in both development and remyelination (Fancy et al. 2009; Ye et al. 2009). The nuclear receptor retinoid X receptor-γ (RXRγ) is a key positive regulator of oligodendrocyte differentiation directly from the analysis of remyelinating tissue (Huang et al. 2011). Electrical activity in demyelinated axons and synaptic signaling to OPCs via AMPA receptors also play an important role in their differentiation (Gautier et al. 2015), while direct electrical stimulation of demyelinated axons can enhance remyelination (Ortiz et al. 2019).

However, differences in the regulation of development and regeneration of myelin do occur; the transcription factor Olig1, although nonessential for developmental myelination (Xin et al. 2005), is required for remyelination where it plays a pivotal permissive role in OPC differentiation (Arnett et al. 2004). In contrast, the Notch signaling pathway, a negative (Wang et al. 1998) or positive (Hu et al. 2003) regulator of differentiation in development (depending on the ligand), is redundant during remyelination, since conditional knockout of the Notch1 gene in OPCs has no or a limited effect on remyelination (Stidworthy et al. 2004; Zhang et al. 2009). The differentiation inhibitory function of endothelin-1 has recently been shown to operate via activation of the Notch pathway, supporting a view that on balance this pathway is inhibitory (Hammond et al. 2014).

Inflammation and Remyelination

The innate immune response to demyelination is important for creating an environment conducive to remyelination (Franklin and Simons 2022). The relationship between inflammation and regeneration is well recognized in many other tissues. However, its involvement in myelin regeneration has been obscured in a field dominated by the immune-mediated pathology of MS and its various animal models such as EAE, where it is unquestionably true that the adaptive immune response mediates tissue damage.

The innate immune component of remyelination involves both CNS resident microglia and monocytes recruited from the circulation, both of which can differentiate into phagocytic macrophages. That the innate immune system plays an important role in remyelination is now firmly established (Kotter et al. 2001; Lloyd and Miron 2019). These roles are many and complex, needing fine tuning so as not to be insufficient to activate a regenerative response nor overly extensive to be the cause of additional damage. The innate immune response during remyelination includes (1) recruitment of inflammatory cells; (2) astrocyte activation; (3) activating and recruiting OPCs and their subsequent differentiation through the expression of a range of growth factors and cytokines; (4) removal of myelin debris that contains inhibitors of OPC differentiation and is required for recycling of cholesterol; (5) extracellular matrix (ECM) production and modification; and, ultimately, (6) resolving inflammation (Franklin and Simons 2022). An activated macrophage phenotype is associated with the recruitment phase of remyelination and the switch to an inflammatory environment dominated by alternatively activated macrophages classically associated with tissue regeneration is causally related to the initiation of differentiation, in part via the production of activin-A (Miron et al. 2013). An area of emerging interest is how microglia register that demyelination has occurred and assume the activation states necessary to contribute to remyelination, with particular focus on the role of pattern-recognition factors such as Toll-like receptors and C-type lectin receptors, and triggering receptors expressed on myeloid cells (TREM2) (Cantoni et al. 2015; Poliani et al. 2015; Cignarella et al. 2020; Cunha et al. 2020; Bosch-Queralt et al. 2021; Gouna et al. 2021). Signaling via these receptors activates microglia via downstream mediators such as nuclear factor-κB (NF-κB) and myeloid differentiation primary response 88 (MyD88) (Cunha et al. 2020). Evidence is now emerging that the adaptive immune response may also contribute to successful remyelination, and in particular regulatory T cells through the production of the ECM-related protein CCN3 (Dombrowski et al. 2017).

DEMYELINATED CNS AXONS CAN ALSO BE REMYELINATED BY SCHWANN CELLS

CNS remyelination can also be mediated by Schwann cells, the myelin-forming cells of the PNS; this occurs in several experimental animal models of demyelination as well as in human demyelinating disease (Snyder et al. 1975; Itoyama et al. 1983, 1985; Dusart et al. 1992; Felts et al. 2005). Schwann cell remyelination occurs preferentially where astrocytes are absent—for example, where they have been killed along with oligodendrocytes by the demyelinating agent (Blakemore 1975; Itoyama et al. 1985). Remyelinating Schwann cells within the CNS were generally thought to migrate into the CNS from PNS sources such as spinal and cranial roots, meningeal fibers, or autonomic nerves following a breach in the glia limitans (Franklin and Blakemore 1993). In support of this idea, CNS Schwann cell remyelination typically occurs in proximity to spinal/cranial nerves or around blood vessels (Snyder et al. 1975; Duncan and Hoffman 1997; Sim et al. 2002a). However, recent genetic fate mapping studies have revealed that very few CNS remyelinating Schwann cells are derived from PNS Schwann cells but instead the majority derive from OPCs (Zawadzka et al. 2010), revealing a remarkable capacity of these cells to differentiate into cells of neural crest lineage as well as other neuroepithelial lineages (astrocytes and oligodendrocytes). This process occurs preferentially in the perivascular niche and is aided by high levels of BMP4 and an absence of astrocyte-derived BMP-antagonist Soctdc1 (Ulanska-Poutanen et al. 2018). Thus, the OPC may be more appropriately considered a broadly multilineage-competent neuroectodermal progenitor capable of producing not only astrocytes and oligodendrocytes, but also Schwann cells, in a context-dependent fashion (Crawford et al. 2014).

The implications of Schwann cell remyelination of CNS axons are unclear (Chen et al. 2021). While both Schwann cell and oligodendrocyte remyelination are associated with a return of saltatory conduction (Smith et al. 1979), their relative abilities to promote axon survival, a major function of myelin (Nave and Trapp 2008), have yet to be established. Thus, from a clinical perspective we do not yet know whether OPC differentiation into Schwann cells has a beneficial or deleterious effect compared to oligodendrocyte remyelination.

CAUSES OF REMYELINATION FAILURE

The efficiency of remyelination is affected by the non-disease-related factors of age and sex (Sim et al. 2002b; Li et al. 2006). These generic factors will have a bearing on the efficiency of remyelination regardless of the disease process involved and will be discussed first.

Like all other regenerative processes, the efficiency of remyelination decreases with age. This manifests as a decrease in the rate at which it occurs and is likely to have a profound bearing on the outcome of a disease process that in the case of MS can occur over many decades. The age-associated effects on remyelination are due to a decrease in the efficiency of both OPC recruitment and differentiation (Sim et al. 2002b). Of these two events, the impairment of differentiation is rate-determining since increasing the provision of OPCs by the overexpression of the OPC mitogen and recruitment factor PDGF following demyelination in old mice does not accelerate remyelination (Woodruff et al. 2004). The impairment of OPC differentiation in aging mirrors the failure of oligodendrocyte differentiation associated with many chronically demyelinated MS plaques (Wolswijk 1998a; Kuhlmann et al. 2008).

The development of protocols for isolating OPCs from aged rodents has allowed a detailed study of the changes that occur in OPCs with aging. In short, OPCs acquire all the hallmarks associated with adult stem cell aging, including mitochondrial dysfunction, loss of DNA repair mechanisms, changes in the epigenome, and dysregulated nutrient signaling (Neumann et al. 2019a). Interestingly, the drivers of OPC aging are not cell intrinsic but rather are determined by features of the niche and especially its biomechanical properties—stiff substrates induce an aged phenotype, while soft substrates induce a youthful phenotype (Segel et al. 2019). The mechanisms governing the OPC age-state involve regulation of the levels of the transcription factor c-Myc, declining levels of which are associated with aging (Neumann et al. 2021). The effects of aging on remyelination are not all cell autonomous as aging will have deleterious consequences for the function of all cell types contributing to the remyelination environment. Best studied are the age-related changes in the innate immune component of remyelination, where aging results in a diminished ability to clear and process myelin debris (Kotter et al. 2006; Natrajan et al. 2015; Cantuti-Castelvetri et al. 2018; Rawji et al. 2018, 2020a).

A key question relating to the development of remyelination therapies is the extent to which age-associated changes can be reversed. This has been demonstrated using the heterochronic parabiosis model—by parabiotic union of a young adult animal to an old adult animal, the old adult animal can be made to remyelinate with the efficiency of a young adult (Ruckh et al. 2012). This is achieved, in part, by the recruitment of circulating young monocytes to bolster the myelin debris clearance function of the old macrophages. Reversing the age-related decline in remyelination efficiency has been achieved by a number of interventions including (1) rejuvenating aged OPCs by disabling their ability to sense a stiff aged niche through the deletion of the mechanoreceptor Piezo-1 or by increasing levels of c-Myc in aged OPCs; (2) by calorie restriction; and (3) by administration of the calorie-restriction mimetic, metformin (Fig. 3; Neumann et al. 2019a,b; Segel et al. 2019).

Figure 3.

Oligodendrocyte progenitor cell (OPC), aging, and rejuvenation. OPCs, like other adult stem cells, undergo functional decline with aging: they have a diminished ability to self-renew and to differentiate. The dysfunction of aged OPCs is underlined by the acquisition of hallmarks of ageing, like DNA damage and mitochondrial dysfunction. The functional capacity of an OPC is determined by its environment (niche) where the physical properties (stiffness) of the brain play a key role in the aging process. Interventions, such as heterochronic parabiosis, activation of the retinoid X receptor-γ (RXRγ), calorie restriction or treatment with metformin, manipulation of substrate (or capacity to sense substrate properties), or partial reprogramming, can reinstate the stem cell potential of OPCs and thereby their capacity for remyelination. Alternatively, aged OPCs can be reprogrammed to a more youthful state. The deletion of Piezo 1 prevents OPCs from sensing the stiffness of the niche. Thus, aged OPCs behave as young OPCs that are normally exposed to a soft environment. Therefore, strategies that restore a more youthful environment or that make OPCs impervious to extracellular changes that occur with ageing lead to a functional rejuvenation of OPCs and thereby store the capacity of aged animals for remyelination. (Figure adapted from Neumann et al. 2019b, and reprinted under the Creative Commons Attribution 4.0 International license [CC BY 4.0 DEED].)

In addition to these generic factors, remyelination could also be incomplete or fail for disease-specific reasons. The strongest evidence for remyelination failure is provided by MS, and the subsequent discussion will specifically relate to this disease, although the issues discussed will be relevant to other diseases with a demyelinating component. Theoretically, remyelination via generation of new oligodendrocytes from OPCs could fail because of (1) a primary deficiency in progenitor cells; (2) a failure of progenitor cell recruitment; or (3) a failure of progenitor cell differentiation and maturation.

Early speculation on remyelination failure focused on the first of these mechanisms, that the process of remyelination itself would deplete an area of CNS of its progenitor cells so that subsequent episodes of demyelination occurring at or around the same site would fail to remyelinate due to a lack of OPCs. However, data from experimental studies indicate that OPCs are remarkably efficient at repopulating regions from which they have been depleted (Chari and Blakemore 2002), and that repeat episodes of focal demyelination in the same area neither depletes OPCs nor prevents subsequent remyelination (Penderis et al. 2003). The situation may be different, however, when the same area of tissue is exposed to a sustained demyelinating insult, where remyelination impairment seems to be due, at least in part, to a deficiency in OPC availability (Mason et al. 2004; Armstrong et al. 2006).

In the second mechanism, MS lesions fail to remyelinate not because of a shortage of available progenitor cells but rather because of a failure of OPC recruitment: proliferation, migration, and repopulation of areas of demyelination. Here, descriptions of demyelinated areas from which oligodendrocyte lineage (OL) cells are absent do indicate that this may account for failure of remyelination in at least a proportion of lesions. Why lesions should become deficient in OPCs is not clear but one possibility is that they are direct targets of the disease process within the lesion. The identification of patients with antibodies recognizing OPC-expressed antigens (NG2) support this possibility (Niehaus et al. 2000). Failure of OPC recruitment into areas of demyelination may arise due to disturbances in the local levels of the OPC migration guidance cues semaphorin 3A and 3F (Williams et al. 2007a).

The best evidence at present supports the third mechanism, a failure of differentiation and maturation, as several sets of observations based on the detection of oligodendrocyte lineage cells within areas of demyelination indicate that this stage of remyelination is most vulnerable to failure in MS. The presence of OPCs apparently unable to differentiate within MS lesions was initially shown with the OL marker O4 (Wolswijk 1998b) and subsequently with NG2 (Chang et al. 2000), PLP (to reveal pre-myelinating oligodendrocytes) (Chang et al. 2002), and with Olig2 and Nkx2.2 (Kuhlmann et al. 2008). Even though the density of OPCs within chronic lesions is on average lower than in normal white matter, the density can be as high as that in normal white matter or remyelinated lesions, showing that OPC availability is not a limiting factor for remyelination.

One possible explanation for this failure of differentiation is that chronically demyelinated lesions contain factors that inhibit progenitor differentiation. First implicated was the Notch–Jagged pathway, a negative regulator of OPC differentiation. Notch and its downstream activator Hes5 were detected in OPCs and Jagged in astrocytes within chronic demyelinated MS lesions (John et al. 2002). However, the expression of Notch by OPCs and Jagged by other cells within lesions undergoing remyelination and, more informatively, the limited remyelination phenotype in experimental models following conditional deletion of Notch in OL cells, suggest that Notch–Jagged signaling is a redundant, nonessential negative regulator of remyelination (Stidworthy et al. 2004).

Other potential inhibitory factors have been identified in different experimental and pathological studies. The accumulation of the glycosaminoglycan inhibitor of OPC differentiation hyaluronan within MS lesions may contribute to an environment within chronic lesions that is not conducive to remyelination by inhibiting OPC function via TLR2 signaling (Back et al. 2005; Sloane et al. 2010). The demyelinated axon itself has been implicated, since demyelinated axons have been shown to express the adhesion molecule PSA-NCAM (Charles et al. 2002), which inhibits myelination in cell culture (Charles et al. 2000). The possibility that the properties of the OPC within areas of demyelination might be regulated by electrical activity (or lack of) in (Gibson et al. 2014) or synaptic input from (Etxeberria et al. 2010) demyelinated axons highlights the complexity of regulatory factors that govern remyelination and by extension account for remyelination failure.

While many studies have concentrated on putative inhibitory signals to account for the failure of OPCs to undergo complete differentiation within demyelinated MS plaques, an alternative explanation is that these lesions fail to remyelinate because of a deficiency of signals that induce differentiation. This hypothesis, based on the absence of factors, is difficult to prove but is consistent with a model of remyelination in which the acute inflammatory events play a key role in progenitor activation and creating an environment conducive to remyelination (see above). While MS lesions are rarely devoid of any inflammatory activity, chronic lesions are relatively noninflammatory compared to the acute lesions and constitute a less active environment in which OPC differentiation might become quiescent. Acute inflammatory lesions are characterized by reactive astrocytes that are the source of many pro-remyelination signaling factors (Williams et al. 2007b; Moore et al. 2011; Molina-Gonzalez and Miron 2019; Rawji et al. 2020b). In contrast, chronic quiescent lesions are characterized by scarring astrocytes that are transcriptionally quiet compared to reactive astrocytes. The scarring astrocyte is better viewed as a consequence of remyelination failure and not its cause. Thus, neither the reactive nor the scarring astrocytes—both contributing to astrogliosis—are likely drivers of remyelination failure.

The two possibilities that remyelination failure reflects the presence of negative factors or the absence of positive factors are not of course mutually exclusive. Moreover, it has become apparent from many studies in recent years that there are a multitude of interacting factors, both environmental and intrinsic, that guide the behavior of OL cells through the various stages of remyelination. Efficient remyelination may depend as much on the precise timing of action as on the presence or absence of these factors. This is called the “dysregulation hypothesis” in which remyelination failure reflects an inappropriate sequence of events (Franklin 2002; Franklin and ffrench-Constant 2008; Fancy et al. 2011). While the causes of remyelination failure in a disease as complex and variable as MS are likely to be many, we still regard this hypothesis as useful for understanding remyelination failure in the majority of cases.

ENHANCING ENDOGENOUS REMYELINATION

Since remyelination can occur completely and the cells responsible are abundant throughout the adult CNS, even within demyelinated lesions, a conceptually attractive approach to enhancing remyelination is to target the endogenous regenerative process (Franklin and Gallo 2014). This approach is predicated on the view that if the mechanisms of remyelination can be understood and non-redundant pathways described, then the causes of remyelination failure and hence plausible therapeutic targets will be identified. From the preceding sections it will be clear that remyelination failure is associated with either insufficient OPC recruitment or, more commonly, failed OPC/oligodendrocyte differentiation. However, the underlying biology of these two phases of remyelination is different and sometimes mutually exclusive. The implication is therefore that pro-recruitment therapies may not promote remyelination where the primary problem is OPC differentiation and vice versa.

A further consideration in the development of remyelination enhancement therapies is the use of appropriate animal models. In the chronic demyelinated plaques of MS, remyelination is assumed to have failed; hence, the requirement is for an intervention that will reactivate a dormant process. In contrast, in many of the demyelination models used to test enhancement of remyelination, such as the toxin-based models, remyelination does not fail. In those models, one can only achieve the acceleration of an already effective ongoing process. That said, assessing the temporal dynamics of remyelination in MS tissue, and whether it has stopped or merely slowed, is difficult to assess from the “snapshots” provided by biopsy or postmortem tissue. Nevertheless, this problem can in part be overcome in two ways. First, by using aged animals in which the slow rate of remyelination is suboptimal presents an opportunity for assessing its enhancement. Second, by modifying standard lesion models in which the endogenous repair process is compromised, such as in the chronic cuprizone model (Armstrong et al. 2006). Theiler's virus-induced demyelination model has also proven useful for demonstrating enhanced remyelination (Njenga et al. 1999). However, assessment of remyelination has proven especially complex in EAE, in which the processes of demyelination and remyelination can occur concurrently. This can make it very difficult to distinguish an effect that renders the environment less hostile to remyelination, allowing it to proceed at its natural rate, from one in which the rate of remyelination is actually accelerated. For example, systemic delivery of putative remyelination-enhancing factors can affect the balance of myelin damage and regeneration via effects on cells other than oligodendroglial cells, such as those of the immune system.

Despite caveats regarding models and methods of analysis, over the last ∼15 years several proof-of-principle studies have allowed the identification of promising pharmacological compounds for enhancing remyelination, both repurposed and newly developed, that modulate key pathways involved in endogenous myelin repair.

Several studies investigating humanized monoclonal antibodies against LINGO-1 (opicinumab) have demonstrated the potential of these compounds to increase axonal myelination both in vitro and in animal models (Mi et al. 2007; Rudick et al. 2008). A promising remyelination potential has also been demonstrated for the antagonists of muscarininc M1 receptor (clemastine, benztropine, quetiapine). In primary OPC cultures, these molecules have been found to be effective in orienting OPCs toward a differentiated phenotype, promoting the formation of mature, myelinating oligodendrocytes (Angelis et al. 2012; Mei et al. 2014; Abiraman et al. 2015; Li et al. 2015). Moreover, in EAE models, a significant improvement in the efficacy of remyelination has been obtained following the removal of the muscarinic acetylcholine receptor M1 in oligodendrocytes (Mei et al. 2016). Several studies have explored the promyelinating potential of modulators of RXRγ (bexarotene, IRX4204), a nuclear receptor with a key role in immune regulation and OPC differentiation and maturation (Huang et al. 2011; Natrajan et al. 2015). The treatment of rodents with a pan-RXR agonist, 9-cis-retinoic acid, following experimental demyelination, resulted in enhanced remyelination, while RXRγ knockout mice showed a significantly reduced number of oligodendrocytes in the lesion area (Huang et al. 2011). Besides RXRγ, other nuclear hormone receptors can be targeted to enhance remyelination, including the vitamin D receptor, the peroxisome proliferator-activated receptor (PPAR), and the liver X receptor (de la Fuente et al. 2015; Veloz et al. 2022). Another class of compounds showing promise as promyelinating treatments are the histamine H3 receptor antagonists (GSK239512, GSK247246), which have been shown to enhance remyelination through increased mature oligodendrocyte differentiation in the cuprizone mouse model (Chen et al. 2017). Several lines of evidence also pointed toward the promyelinating potential of steroid hormone-based therapies (androgens, estrogen, thyroid hormones analogs), which have been shown to enhance myelin repair, possibly via the induction of OPC differentiation (Sutiwisesak et al. 2021). Furthermore, steroids (clobetasol) and antifungals (miconazole) have also demonstrated notable remyelinating properties in experimental models, although results have not been completely consistent across models. Both clobetasol and miconazole have been found to promote oligodendrocyte differentiation and increase remyelination in a rodent model of demyelination (Najm et al. 2015). Interestingly, many of these compounds (clemastine, benzatropine, tamoxifen, bazedoxifene, and miconazole), while targeting completely different pathways, all share the property of promoting oligodendrocyte maturation through the same signaling mechanism, which is the increased production of 8,9-unsaturated sterols (Hubler et al. 2018).

Of note are compounds that combine an anti-inflammatory with a remyelinating mechanism of action, such as siponimod, a modulator of the sphingosine-1 receptor (S1P-R) that crosses the blood–brain barrier and selectively targets S1P1-R and S1P5-R. While significantly reducing lymphocyte migration in the CNS through S1PR1, siponimod may also promote remyelination via S1PR5 expressed on oligodendrocytes (Behrangi et al. 2022).

Despite the significant number of molecules known to have promyelinating properties, so far, only a limited subgroup has advanced to being tested in phase 2 clinical trials in patients with MS. While most of the already completed and ongoing clinical trials testing promyelinating drugs are conducted in patients with acute or chronic optic neuritis, and employ the change in the latency of visual evoked potentials (VEPs) as an outcome measure (with a reduced VEP latency interpreted as reflecting remyelination), a few of them investigate patients with relapsing MS, with an either relapsing-remitting or progressive disease course (Lubetzki et al. 2020). In this case, changes in lesion MTR, with increasing MTR values over time being interpreted as possibly reflecting remyelination, or changes in multicomponent scores of clinical function, are employed as outcome measures.

While preclinical studies identified opicinumab, a monoclonal antibody directed against LINGO-1, as a promising molecule, the results of clinical trials testing this molecule in patients with MS have been disappointing (Cadavid et al. 2017, 2019). The safety and efficacy of the retinoid receptor agonist bexarotene have been tested in the clinical trial CCMR One, in which 52 patients with RRMS received either 300 mg/m² of BSA per day of oral bexarotene or oral placebo for 6 months. In this study, the change in mean lesional MTR over the 6-month follow-up was employed as the primary efficacy outcome (Brown et al. 2021). While bexarotene was found to be associated with an increased number of adverse events, the change in mean lesional MTR did not significantly differ between bexarotene and placebo. Nevertheless, this clinical trial demonstrated that MS lesions are heterogeneous in their capacity to remyelinate following treatment with bexarotene. In particular, lesions that were more demyelinated at baseline as well as lesions that were localized in the gray matter showed greater remyelination following this treatment. Although the results of this study do not support the clinical use of bexarotene due to its poor tolerability, future clinical trials may investigate the remyelinating properties of other RXR agonists with better tolerability profiles.

Among the few clinical trials testing promyelinating treatments that have been already completed, there is also ReBUILD, a double-blind, crossover trial, which compared a group of patients with chronic optic neuritis on the antagonist of the muscarinic M1 receptor clemastine (5·36 mg orally twice daily) for 90 days followed by placebo for 60 days, with a group of patients on placebo for 90 days followed by clemastine (5·36 mg orally twice daily) for 60 days (Green et al. 2017). In this study, clemastine was shown to reduce the VEP latency delay without being associated with serious adverse events. Based on the promising results of ReBUILD, clemastine is currently being tested in the context of acute optic neuritis (NCT02521311) and as combinatory remyelinating therapy (NCT03109288).

The histamine H3 receptor blocker GSK239512 has also been tested in a phase II clinical trial in patients with relapsing-remitting MS, showing a limited yet significant effect in increasing the MTR value of white matter lesions in treated patients compared to patients on placebo (Schwartzbach et al. 2017). The results of two clinical trials exploring the remyelinating properties of nanocrystalline gold (NCT03536559) and of the estrogenic compound bazedoxifene (NCT04002934), which have been tested in patients with MS chronic optic neuritis, are expected very soon. Finally, the remyelinating effects of testosterone in relapsing-remitting MS are being tested in a phase II clinical trial that is currently recruiting (NCT03910738).

In addition to the opportunities provided by the pharmacological manipulation of different pathways involved in remyelination, there is another intriguing therapeutic strategy to enhance remyelination, based on the growing evidence suggesting that axonal electrical activity has a strong influence on oligodendrogliogenesis and (re)myelination. In particular, silencing electrical activity using tetrodotoxin has been demonstrated to reduce the percentage of myelinated fibers, in vitro in myelinating cocultures and in vivo in the optic nerve after intravitreal injection (Demerens et al. 1996). Conversely, through the optogenetic stimulation of neurons of the murine premotor cortex, it has been demonstrated that neuronal activity promotes the proliferation of OPCs, as well as oligodendrocyte maturation and myelination, with functional improvement (Gibson et al. 2014). In demyelinating conditions, axonal electrical activity remains essential to promote remyelination. Indeed, it has been demonstrated that, after demyelination, axons still have the potential to conduct electrical impulses. This neuronal activity regulates OPC differentiation by synaptic release of glutamate. Blocking action potentials with tetrodotoxin leads to increased OPC apoptosis and failure of OPC differentiation into mature oligodendrocytes (Gautier et al. 2015). Based on this evidence, electrical stimulation is currently being considered as a promising therapeutic approach to enhance myelin repair in demyelinating conditions. In this perspective, transorbital electrical stimulation is currently being tested as a means to shorten VEP latency in MS patients with acute optic neuritis (NCT04042363).

REMYELINATION BY GLIAL PROGENITOR CELL TRANSPLANTATION

The mobilization of endogenous oligodendrocyte progenitor cells may be appropriate for a variety of disorders of progressive dys- or demyelination. However, a broad swath of neurological disease involves the structural loss of white matter, reflecting frank loss of both oligodendrocytes and their progenitors, and often of their associated fibrous astrocytes. These disorders include ischemic and traumatic demyelination, as well as demyelination associated with sustained autoimmune inflammation, and insults as diverse as chemotherapy and radiotherapy. In addition, many of the neurodegenerative disorders—including Huntington's disease, Alzheimer's disease, and schizophrenia, among others—are attended by progressive demyelination, as is nominally healthy aging itself. Moreover, the childhood hereditary and metabolic disorders of white matter, the hypomyelinating leukodystrophies and lysosomal storage disorders in particular, are intrinsically refractory to any approach aimed at activating resident progenitors—which themselves carry the culpable mutation. Together, these diverse disorders of myelin require extensive tissue repair, and in many cases even whole neuraxis myelination. Thus, any practical cell therapeutics for the myelin disorders must provide large numbers of progenitors biased to oligodendrocyte differentiation and myelinogenesis.

This requirement prompted the development of human OPCs from a variety of sources, including human tissue (Roy et al. 1999; Dietrich et al. 2002; Windrem et al. 2002, 2004) and later from embryonic and induced pluripotent stem cells (Izrael et al. 2007; Hu et al. 2009; Stacpoole et al. 2013; Wang et al. 2013; Douvaras et al. 2014; Piao et al. 2015). Each of these sources has now been shown able to generate potentially myelinogenic oligodendrocytes (Fig. 4), and each has its own strengths and weaknesses as a cellular therapeutic (Goldman et al. 2012, 2021). Of note, OPCs may also be generated directly from mouse and human fibroblasts as well (Najm et al. 2013; Yang et al. 2013; Ehrlich et al. 2017; Tanabe et al. 2022), suggesting the potential development of yet another source of clinically appropriate OPCs (Goldman 2013).

Figure 4.

Glial progenitor cell (GPC) sources, phenotypes, and clinical targets. GPCs may be directly sorted from tissue or produced from either human embryonic stem cells (hESCs) or human-induced pluripotential cells (hiPSCs) and then immunoselected based on their expression of either gangliosides recognized by monoclonal antibody (mAb) A2B5 or of CD140a/PDGFαR. The CD140a phenotype includes all potential oligodendrocytes, while the tetraspanin CD9 and sulfatide-directed mAb O4 identify progressively more oligodendrocyte-biased fractions (Sim et al. 2011; Douvaras et al. 2014). In contrast, CD44 recognizes a more astrocyte-biased fraction (Liu et al. 2004). The choice of tissue-, hESC-, or iPSC-derived GPCs depends upon whether allogeneic or autologous grafts are desired. Whereas autologous grafts of iPSC-derived GPCs might obviate the need for immunosuppression, their generation may take months, and their use in the hereditary leukodystrophies would first require correction of the underlying genetic disorder in the donor cell pool; at present, such genetic disorders of myelin would be better approached with allografted tissue- or hESC-derived GPCs. (This figure is adapted and reprinted, with permission, from Goldman et al. 2012.)

To serve as practical, safe, and effective therapeutic vectors, OPCs must be deliverable in both reliable purity and significant quantity. To address this need, several methods for isolating human OPCs (hOPCs) from mixed cell populations have been developed, which use fluorescence activated or magnetic bead-based cell sorting targeting OPC surface antigens (Roy et al. 1999; Nunes et al. 2003; Sim et al. 2011). These hOPC isolates were first obtained from human brain tissue, and their ability to robustly remyelinate demyelinated tissue was initially established in animal models of both congenital and adult dysmyelination (Windrem et al. 2002, 2004). In Shiverer mice, a dysmyelinated mouse deficient in Myelin basic protein, fetal and adult tissue-derived hOPCs were noted to behave differently after neonatal xenograft. In particular, isolates derived from adult white matter myelinated recipient brain much more rapidly than fetal hOPCs; adult-derived progenitors achieved widespread myelination just 4 weeks after graft, while cells derived from second trimester fetuses took over 3 months to do so. On the other hand, these fetal hOPCs expanded in vivo, emigrated more widely, engrafted more efficiently, and differentiated as both astrocytes and oligodendrocytes, all suggesting their greater potential as therapeutic vectors (Fig. 5; Windrem et al. 2004; Sim et al. 2011). Nonetheless, despite the mitotic competence of these cells, they remain finite in both initial number and in expansion competence, necessitating their periodic reacquisition from new donors, and thus rendering their routine use impracticable. Recent efforts have thus focused on the generation of myelinogenic OPCs from pluripotent human stem cells as preferred sources for the scalable development of therapeutic hOPCs.

Figure 5.

Transplanted fetal human glial progenitor cells (hGPCs) myelinate the congenitally unmyelinated shiverer brain. (A–E) Myelination of congenitally hypomyelinated Shiverer (Shi) mice by human fetal tissue-derived glial progenitor cells. (A,B) Representative sagittal images of an engrafted Shi/Shi × Rag2^−/− brain, sacrificed at 1 yr of age. (A) Human donor cells identified by an antihuman nuclear antibody (hN; red). (B) Donor-derived myelin basic protein (MBP; green) in sections adjacent or nearly so to matched sections in A. All major white matter tracts heavily express MBP (which is all donor-derived in Mbp-null Shiverer mice). (C) Sagittal view through cerebellum of a year-old engrafted Shi/Shi × Rag2^−/− brain. All cells were stained with DAPI (blue); donor cells were identified by human nuclear antigen (hN, red) and donor-derived myelin by MBP (green). (D) Reconstituted nodes of Ranvier in the cervical spinal cord of a transplanted and rescued 1-yr-old shi/shi × rag2^−/− mouse, showing paranodal Caspr protein and juxtaparanodal potassium channel Kv1.2, symmetrically flanking each node. Untransplanted Shiverer brains do not have organized nodes of Ranvier and, hence, cannot support saltatory conduction (Caspr, red; Kv1.2, green). (E) Electron micrograph of a 16-wk-old Shiverer mouse implanted perinatally with hGPCs shows a Shiverer axon with a densely compacted myelin sheath. Scale bars, 5 µm (D); 1 µm (E). (Images A and B from Windrem et al. 2008, C and D from Goldman et al. 2012, and E from Windrem et al. 2004, reprinted with permission. This figure is adapted, with permission, from Osorio and Goldman 2016 and Mariani et al. 2019.)

PLURIPOTENT STEM CELLS AS A SOURCE OF MYELINOGENIC PROGENITORS

Following the first report of myelination in the injured spinal cord by implanted murine embryonic stem cells (mESCs) (Brüstle et al. 1999), oligodendrocytes derived from human ESCs (hESCs) were reported to similarly myelinate demyelinated foci in spinal cord contusions (Nistor et al. 2005). hESC-derived hOPCs were then shown to generate myelin in the rodent brain (Zhang et al. 2001), and subsequent studies with improved differentiation protocols reported more efficient glial and oligodendrocytic differentiation (Izrael et al. 2007; Hu et al. 2009). Nonetheless, none of these studies isolated hOPCs prior to transplant, nor did any follow animals for the periods of time required to ensure the stability and safety of the engrafted cells. This was problematic, since persistent undifferentiated hES cells in the donor pool may retain the potential to generate teratomas or neuroepithelial tumors (Roy et al. 2006; Pruszak et al. 2009). To address these considerations, Wang et al. (2013) developed a more stringent protocol for differentiating hOPCs from pluripotent human stem cells, which yielded highly enriched and myelinogenic populations of hOPCs, with no detectable residual stem cells nor evidence of tumorigenicity after transplant. In particular, Wang and colleagues noted that the risk of tumorigenesis and unintended expansion from hESCs could be effectively mitigated by extended differentiation protocols and rigorous quality assessments during cell production. Nonetheless, hESC lines are derived from human embryos, which has hindered their use in a number of both political jurisdictions and cultures. Enthusiasm has thus developed for the potential use of induced pluripotent stem cells (iPSCs) as a more universally acceptable and clinically scalable source of OPCs.

iPSCs are pluripotent cells that have been generated by the reprogramming of somatic cells to a less phenotypically committed stem cell ground state, by the forced expression of transcription factors that permit the self-renewing stem cell phenotype (Stadtfeld and Hochedlinger 2010). Most typically, iPSCs are generated from somatic cells cotransduced with a set of stem cell–associated transcription factors, including POU5F1/OCT4, SOX2, MYC, KLF4, and NANOG (Yamanaka 2007, 2008), or alternatively, with miRNAs that modulate the expression of their encoded proteins, or small molecules that mimic their actions (Lin et al. 2009; Anokye-Danso et al. 2011). iPSCs have the advantage over hESCs of yielding lineage-restricted cells that can be autologously transplanted back to the subjects from which they were generated, potentially obviating the need for posttransplant immune suppression. The differentiation strategies for generating hOPCs from hESCs have proven adaptable to iPSCs as well, and yield both glial progenitors and terminally differentiated oligodendrocytes that are highly myelinogenic in vivo (Czepiel et al. 2011). More recent advances in optimizing the methods for generating hOPCs from human iPSCs and ESCs have led to the production of highly enriched populations of hOPCs that efficiently myelinate in vivo, while manifesting no evidence of tumorigenic potential (Fig. 6; Wang et al. 2013; Douvaras et al. 2014; Piao et al. 2015).

Figure 6.

Human pluripotent stem cell–derived glial progenitor cells (GPCs) can remyelinate dysmyelinated hosts. (A) This schematic outlines the multistage protocol by which GPCs, oligodendrocytes, and astrocytes may be generated from human pluripotent stem cells, whether human embryonic stem cells (hESCs) or induced pluripotent stem cells (iPSCs). (B–F) Representative images taken at serial stages of glial differentiation, with the serial expression of selected marker proteins noted at each stage. (G) 3 mo after neonatal transplant into hypomyelinated Shiverer mice, human-induced PSC (hiPSC)-derived GPCs have matured as myelinating, myelin basic protein (MBP)-expressing oligodendrocytes (MBP, green; human nuclear antigen, red). (H) The hiPSC–derived oligodendrocytes ensheath mouse axons (neurofilament, red; MBP, green). (I) hiPSC-derived oligodendrocytes can myelinate the entire brain of Shiverer mice, which do not otherwise express MBP (green). (J) The myelin generated by hiPSC oligodendrocytes is ultrastructurally normal, exhibiting major dense lines and thick myelin sheaths. The use of such serial and distinct stages of growth factor exposure, paired with more extended periods of differentiation, have led to the production of highly enriched populations of human GPCs (hGPCs), that are highly efficient at myelinogenesis in vivo while manifesting no evident tumorigenesis. Scale bars, 100 µm (B–E); 25 µm (F); 100 µm (G); 10 µm (H); 100 nm (J). (This figure is adapted, with permission, from Wang et al. 2013 and Mariani et al. 2019.)

The capability to now produce scalable quantities of highly myelinogenic iPSC-derived hOPCs allows us to realistically consider their use for treatment of myelin disorders. These iPSC hOPCs have the potential advantage of autologous use, so as to avoid the immunogenicity of OPCs derived from an otherwise unmatched allogeneic line. Such an autologous strategy, while not readily scalable to larger patient populations, may nonetheless be appropriate for individual cases of vascular, traumatic, and inflammatory demyelination. In this latter set of disorders, no underlying genetic lesion may exist, and sufficient axonal numbers may remain for remyelination to be beneficial. Indeed, after genetic correction using contemporary genome editing (Gaj et al. 2013), hOPCs may be generated from edited iPSCs derived from patients harboring pathological mutations, and then transplanted back into those same patients. Such genetically corrected autologous grafts may permit the treatment of genetic disorders of myelin with curative intent, yet without the need for immune suppression.

CHALLENGES IN THE USE OF PLURIPOTENT STEM CELL–DERIVED hOPCS

The autologous transplantation of iPSC-derived hOPCs generated from individual patients is an attractive proposition, but in practice one that would be difficult to scale up to large patient populations. As such, whether OPCs are derived from hESCs or iPSCs, it is likely that they will need to be delivered as allogeneic grafts to genetically unmatched recipients. Because of the likelihood of rejection of incompatible immunophenotypes, these strategies may thus be limited by a need for long-term immunosuppression in graft recipients. Yet chronic immunosuppression carries its own significant risks and morbidities, suggesting the need for developing donor cells able to avoid postgraft immunodetection and rejection. To this end, a number of laboratories have developed strategies for producing hypoimmune pluripotent stem cell lines whose derivatives might avoid immune recognition, whether by HLA class 1 and 2 antigen deletion, substitution, or knockdown (Gornalusse et al. 2017; Deuse et al. 2019; Xu et al. 2019), or the expression of checkpoint inhibitors such as PDL1 or CD47 (Gornalusse et al. 2017; Deuse et al. 2019), or combinations thereof (Lanza et al. 2019; Malik et al. 2019). Most recently, the concurrent use of CD47 overexpression as a means of suppressing dendritic cell activity, while knocking out HLA class 1 and 2 antigens by the concurrent knockout of β2 microglobulin and CIITA, has emerged as a compelling strategy for establishing immunoavoidance. Somatic cells derived from iPSCs edited by this approach have durably engrafted in both blood and solid organs, after both allogeneic and xenogeneic transplantation (Hu et al. 2023a,b,c).

These “off-the-shelf” universal hypoimmune pluripotent stem cells may avoid or minimize the need for host immunosuppression, and have the added advantage of reducing treatment cost, as single master cell banks might be employed to produce a range of desired target cell phenotypes. That said, none of these strategies has yet been studied well in the CNS, and so none have yet been shown to prevent the immune rejection of cells transplanted into the adult brain or spinal cord. Furthermore, the attractiveness and potential utility of these second-generation approaches to immunoavoidance are not without risk; all carry the risk of enabling expansion or unintended differentiation that cannot be checked by the immune system, should tumorigenic or otherwise undesired cell types arise after transplant. As a result, some means of conditionally ablating undesired donor cells might be prudent, as might engineering the cells to undergo induced death upon pharmacological direction (Liang et al. 2018). A number of such strategies are now under development and appear promising (Jones et al. 2014; Sheikh et al. 2021), although the long-term safety and efficacy of each still needs to be demonstrated in the CNS. In that regard, while initial studies using an inducible Caspase9 system proved successful in eliminating undifferentiated cells from NSC grafts (Itakura et al. 2017), the brain presents a number of unique issues to the use of such a “suicide switch” strategy. These concerns include both the blood–brain barrier permeability of prodrugs needed to trigger donor cell death and the potential for cerebritis and cerebral edema in the setting of targeted cell clearance; the latter presents the possibility of an immune effector cell-associated neurotoxicity syndrome (ICANS)-like morbidity (Siegler and Kenderian 2020; Sheth and Gauthier 2021). Thus, while the development of immunoavoidant lines and targeted donor cell elimination strategies may be crucial to the development of safe and effective clinical vectors, neither is without risks that will need to be considered carefully in patient selection and treatment.

Aside from their pluripotency and the risks as well as opportunities that this may entail, iPSCs may retain some of the epigenetic marks—the DNA methylation and histone acetylation patterns of chromosomal architecture—of the donor cells from which they are derived (Stadtfeld et al. 2010). As a result, their age and cell type of origin may influence the differentiation competencies of iPS cells derived from different tissue sources and from subjects of different ages (Polo et al. 2010). This is of potential concern in that if iPSCs do not undergo complete reprogramming, then their derived hOPCs might be expected to differ from their tissue or hESC-derived homologs (Kim et al. 2010; Polo et al. 2010). Whether any such differences will prove meaningful in clinical practice remains to be seen, yet it would seem most wise to pursue the clinical use of hOPCs produced from iPSCs derived from the youngest subjects possible, with RNA expression patterns, DNA and histone methylation marks, and chromatin occupancies all as close to those of naive hES cells as possible.

It is worth noting here that hOPCs may be generated not only from pluripotent cells, but directly from somatic cells, using phenotype-defining transcription factors. As noted, several recent studies have reported the direct induction of both OPCs and oligodendrocytes from fibroblasts, using targeted overexpression of defined proglial transcripts (Najm et al. 2013; Yang et al. 2013; Ehrlich et al. 2017; Tanabe et al. 2022). Such avoidance of pluripotent intermediates in the generation of glial progenitors may mitigate the risk of tumorigenesis; yet the mitotic potential, in vivo differentiation efficiencies, and durabilities of such directly induced OPCs and their derived oligodendrocytes still need to be established. Indeed, the phenotypic stability of such cells remains unclear, since at least some lineages generated via direct conversion may retain the ability, and perhaps a proclivity, to revert back to their parental phenotype. For instance, Scholer and colleagues (Kim et al. 2021) demonstrated the metastability of OPCs directly induced from pericytes, which reverted back to pericyte lineage after in vivo transplantation and in vivo residence. The extent to which such phenotypic instability is a function of time, source phenotype, reprogramming strategy employed, and host environment all remain to be established, and will need to be done before OPCs generated via direct induction might be considered as potential clinical vectors. That said, the already available panoply of human pluripotent stem cells, both embryonic and induced, and the potential use of directly reprogrammed cells as well, augur well for the availability of scalable sources of human OPCs going forward.

PEDIATRIC TARGETS OF PROGENITOR CELL–BASED THERAPIES FOR MYELINOGENESIS

Tens of thousands of children in the United States suffer from diseases of myelin failure or loss. These include the metabolic demyelinations such as adrenoleukodystrophy; the lysosomal storage disorders, such as metachromatic leukodystrophy, neuronal ceroid lipofuscinoses, mucopolysaccharidoses and gangliosidoses, Niemann–Pick disease, and Krabbe's disease; the hypomyelinating diseases, such as Pelizaeus–Merzbacher disease and hereditary spastic paraplegia; the myelinoclastic disorders, vanishing white matter disease and Canavan's disease; dysmyelinating disorders such as Alexander's disease (Brenner et al. 2001; Dietrich et al. 2005; Li et al. 2005; Bugiani et al. 2011); and most commonly of all, periventricular leukomalacia, the most common form of cerebral palsy (Back and Rivkees 2004; Follett et al. 2004; Robinson et al. 2005). Their mechanistic heterogeneity notwithstanding, all of these conditions include the prominent loss of oligodendrocytes and central myelin, highlighting their potential attractiveness as targets for cell replacement (Helman et al. 2015; Goldman et al. 2021).

Neonatal Delivery of OPCs for Enzymatic Reconstitution and Myelin Preservation

Since OPC engraftment is both widespread and associated with astrocytic as well as oligodendrocytic production, glial progenitors would seem an especially promising vehicle for dispersing functionally competent glia throughout otherwise diseased and/or enzyme-deficient brain parenchyma. The lysosomal storage disorders present especially attractive targets in this regard (Snyder et al. 1995), since wild-type lysosomal enzymes may be released by integrated donor cells, and taken up by deficient host cells through the mannose-6-phosphate receptor pathway (Urayama et al. 2004). As a result, a relatively small number of donor glia may provide sufficient enzymatic activity to correct the underlying catalytic deficit and storage disorder of a much larger number of host cells (Lacorazza et al. 1996; Jeyakumar et al. 2005). The intracerebral delivery of OPCs would thus seem an especially attractive approach for treating those demyelinating diseases associated with enzyme deficiencies specific to brain. By way of example, metachromatic leukodystrophy (MLD) is characterized by deficient expression of Arylsulfatase A (ARSA), which results in sulfatide misaccumulation and oligodendrocyte loss. Mesenchymal and hematopoietic stem cell grafts have proven unable to correct the CNS manifestations of this disorder (Koç et al. 2002). In contrast, experimental models of MLD have responded well to OPC grafts (Givogri et al. 2006), suggesting that the broader dispersal competence and greater histiotypic appropriateness of orthotopically engrafted OPCs might provide significant therapeutic advantages relative to nonneural phenotypes. Similarly, while asymptomatic Krabbe patients transplanted with umbilical cord stem cells manifested slower disease progression than untreated controls, the benefits of transplantation to children engrafted after symptom onset have been limited (Escolar et al. 2005). Yet the intracerebral parenchymal infiltration of stromal derivatives is minimal, suggesting that treatment of these children with native, brain-penetrant NSCs or glial progenitors might comprise a more promising treatment strategy (Pellegatta et al. 2006). More broadly, these early efforts speak to the potential of engrafted NSCs and OPCs as vehicles for intracerebral enzyme replacement, in the lysosomal storage disorders as well as in the broader category of disorders of brain metabolism with associated dysmyelination.

Neonatal Delivery of OPCs for Structural Myelin Replacement

To assess the potential of OPC-based treatment for congenital dysmyelination, Windrem and colleagues transplanted sorted hOPCs of both fetal and adult origin into newborn hypomyelinated Shiverer mice (Windrem et al. 2002, 2004). This work followed similar studies in which both native and immortalized murine NSCs had been transplanted into Shiverers; each yielded some degree of context-dependent differentiation and myelination (Yandava et al. 1999; Mitome et al. 2001). On that basis, Windrem and colleagues extracted fetal hOPCs from second-trimester forebrain as well as adult OPCs from surgically resected subcortical white matter; both were isolated by either fluorescence-activated or immunomagnetic sorting, based on the phenotype A2B5⁺/PSA-NCAM^–. When transplanted into recipient Shiverer mice, both fetal and adult hOPCs spread widely throughout the brain and developed as both astrocytes and oligodendrocytes (Fig. 5). The cells did so in a highly context-dependent fashion, such that those donor cells that engrafted presumptive white matter developed as myelinogenic oligodendrocytes and fibrous astrocytes, while those invading cortical and subcortical gray either remained as progenitors or differentiated largely as protoplasmic astrocytes (Windrem et al. 2004).

Following a single neonatal intracallosal injection, the majority of donor cells engrafted the presumptive white matter, such that the corpus callosum and capsules densely expressed myelin basic protein throughout the forebrain white matter tracts (Windrem et al. 2004). Donor-derived myelin effectively ensheathed host Shiverer axons, as validated by both confocal imaging and by the ultrastructural observation of donor-derived compact myelin—of which native Shiverer oligodendrocytes are incapable. Confocal analysis also revealed nodes of Ranvier between donor-derived myelinated segments, while transcallosal conduction velocities were normalized in the OPC-transplanted mice. Using a multisite delivery approach intended to achieve broader cell dispersal in the recipient CNS, Windrem et al. (2008) then observed cell colonization throughout the entire neuraxis, with effective whole-neuraxis myelination that included the entire brain, brainstem, cerebellum, and cranial nerve roots, along with much of the spinal cord and roots. This was associated with significantly and substantially prolonged survivals in transplanted Shiverer mice, with frank rescue and phenotypic recovery of a large minority (Windrem et al. 2008); whereas untreated Shiverers invariably die by 20 weeks of age, some engrafted animals achieved normal murine life spans. These data strongly suggested the feasibility of neonatal OPC implantation as a strategy for treating the congenital disorders of myelin formation and maintenance. Later studies refined the criteria for selecting myelinogenic progenitors, by identifying the PDGFa receptor epitope CD140a as recognizing the population of potentially oligodendrocytic cells (Sim et al. 2011). OPCs selected on the basis of CD140a expression proved superior to those selected on the basis of A2B5 in their efficiency, rapidity, and fidelity of differentiation and ultimate extent of myelination; as such, they have supplanted the latter as a preferred cellular vector for therapeutic remyelination.

Challenges in the Use of Glial Progenitor Grafts for Childhood Myelin Disorders

Cell-based therapies comprise a broad platform for the potential amelioration of enzymatic and storage disorders, in that a common strategy of intracerebral delivery of OPCs may prove broadly applicable across a variety of specific enzymatic disorders. In practice though, given treatment regimens will need to be tightly calibrated to specific disease phenotypes and stages. Little data are available as to the number or proportion of wild-type cells required to achieve local correction of enzymatic activity and substrate clearance in any storage disorder, and these values will likely need to be empirically derived for different disease targets. Similarly, the efficiency and extent of myelination required to achieve significant benefit in hypomyelinating disorders remains unclear, and will depend upon disease extent and duration as much as donor cell dispersal and myelinogenic competence in the disease environment. These caveats notwithstanding, there is considerable reason for optimism that cell-based therapy of the pediatric myelin disorders—including not only the storage disorders, but also the primary dysmyelinations such as Pelizaeus–Merzbacher disease (Osorio et al. 2017; Osório and Goldman 2018) vanishing white matter disease (Bugiani et al. 2011; Dooves et al. 2019), and other hypomyelinating leukodystrophies (Helman et al. 2015; Wolf et al. 2021) may soon prove feasible (Goldman 2011). In that regard, a phase 1 trial of implanted NSCs reported the long-term safety of NSC implants into children with Pelizaeus–Merzbacher disease, with some evidence on imaging of new myelin formation at the sites of implantation (Gupta et al. 2012, 2019). One may expect that with the development of improved and scalable preparations of lineage-restricted hGPCs, that this treatment strategy may offer tangible benefit to these especially needy pediatric patient populations.

Adult Disease Targets of Glial Progenitor Cell–Based Treatment

In adults, oligodendrocytic loss is causally involved in diseases as diverse as traumatic brain injury, MS and its variants, and hypertensive and diabetic white matter loss. In addition, the prominent role of oligodendrocytic loss in both vascular dementia and age-related white matter loss is becoming increasingly recognized (Kalaria 2018). All of these disorders are potential targets of glial progenitor cell replacement therapy, although the adult disease environment may limit the feasibility of this approach in ways not encountered in pediatric targets (Franklin and ffrench-Constant 2008). For instance, the chronically ischemic brain tissue of diabetic and hypertensive patients with small vessel disease may present inhospitable environments for graft acceptance, and require aggressive treatment of the underlying vascular disorders before any cell replacement strategy may be considered. Similarly, the inflammatory disease environment of patients with autoimmune demyelination presents its own challenges, which need to be overcome before cell-based remyelination can succeed. Nonetheless, current disease-modifying strategies for treating both vascular and autoimmune diseases have advanced to the point where transplant-based remyelination of adult targets may now be feasible (Goldman et al. 2012, 2021).

Multiple Sclerosis and Autoimmune Demyelination

Most experimental models of cell-based remyelination have focused on MS, a debilitating disease characterized by both inflammatory myelinolysis and degenerative axonal loss. The attraction of MS as a therapeutic target derives from its high incidence and extraordinary prevalence, given its typical onset in youth and long disease course. In the United States alone, over 200 young adults are diagnosed weekly, with more than 300,000 affected nationally. Within 10 years, roughly one-half of MS patients develop the progressive neurodegeneration of secondary progressive MS. In the past, MS had not been an actively researched target for cell therapy, since there was limited enthusiasm for introducing new cells into an active disease environment—one essentially primed for allograft rejection. Yet the continuous improvement of approaches toward CNS immune modulation have so substantially diminished disease recurrence, as to make the use of cell replacement strategies for restoring myelin to demyelinated lesions more tenable.

In that regard both NSCs and OPCs have been assessed as potential cell therapeutics for myelin repair, in a variety of models of acquired demyelination (Pluchino et al. 2004; Franklin and Kotter 2008). As cellular vectors for remyelination, NSCs have thus far disappointed in preclinical studies, as their efficiency of oligodendrocytic differentiation and myelination appears low. In contrast, the intracerebral delivery of OPCs into demyelinated brain has been associated with substantial oligodendrocytic differentiation and myelination in a variety of models of adult demyelination (Windrem et al. 2002). Most notably (Duncan et al. 2009), fetal tissue–derived hOPCs transplanted into cuprizone-demyelinated adult mouse brain have been shown to effectively remyelinate host axons, with both neurological improvement and rescue of callosal conduction velocity (Fig. 7; Windrem et al. 2020), while PSC-derived OPCs have been shown to effectively remyelinate demyelinated tissue in rodent models of both spinal cord injury (Kawabata et al. 2016) and stroke (Martinez-Curiel et al. 2023). Thus, when provided a permissive axonal substrate, donor hOPCs serve as effective cellular vectors for remyelination.

Figure 7.

Myelination by adult-transplanted human glial progenitor cells (hGPCs) restores function in Shiverer mice. (A) Mice were injected bilaterally in the corpus callosum and striatum at 4 wk of age with either cells or vehicle. When 18 wk old, the mice were examined sequentially in a series of tests spanning behavior to anatomy. The arrows indicate the sequence of tests and italics indicate the relevant figures. (B) Sham-injected (top) and hGPC-transplanted (bottom) Shiverers were videoed from below on a treadmill, using DigiGait. (C) Of the eight Shiverer controls and eight transplanted Shiverers tested on the treadmill, only two of eight untreated mice were able to walk on the treadmill, while seven of eight engrafted mice did so on their first try (P = 0.012, chi-squared), and all eight did so on a second attempt. (D) Schematic of the measurements taken to assess the trans-callosal response to electrical stimulation of sampled slices. (E) Representative traces of normally myelinated shiverer heterozygous mice (wild-type [WT]), sham-injected Shiverers (sham), and transplanted Shiverers (Tpt) demonstrating restoration of the fast conduction N1 component in the transplanted mice. (F) N1 amplitude shows no significant difference between slice preparations derived from normal WT and transplanted (Tpt) Shiverer brains, suggesting transplant-mediated normalization of the ratio of myelinated to unmyelinated axons in the transplanted Shiverer brain. Untreated Shiverers have effective N1 velocities of zero, and cannot be assessed as such. (G) The slow component, N2 conduction velocity of transplanted callosal slices was significantly more rapid than that of untreated Shiverers, and no different from that of WT. (H–P´) All images are taken from 18-wk-old Shiverers injected intracallosally at 4 wk with hGPCs. (H) Low-magnification image of the corpus callosum of one hemisphere of an adult-engrafted Shiverer. hN (red), (myelin basic protein [MBP]; green). (I) Human oligodendrocytes, coimmunostained for human cytoplasmic antigen (hCyto) and MBP; the latter is necessarily human, as Shiverer mice do not make MBP. (J) A single mature donor-derived oligodendrocyte. Right-hand images show respective color splits for human cytoplasm (red) and MBP (green), showing the many myelinating sheaths produced by a single hGPC-derived oligodendrocyte. (K–L). Electron micrographs (EMs) of sham control corpus callosum demonstrating the cytoplasmic inclusions in the myelin sheath characteristic of Shiverers. (M–P´) EMs of transplanted shiverers. The insets show the major dense lines (arrowheads), characteristic of compact myelin. Scale bars, 100 µm (H); 20 µm (I,J); 500 nm (K,L); 200 nm (M–P); 100 nm (M´–P´). (This figure is adapted from Windrem et al. 2020 and reprinted under the terms of the Creative Commons CC-BY license.)

Challenges in the Use of OPC Transplantation for Adult Disorders

Regardless of the cell-autonomous capabilities of donor OPCs, the complexity of the adult disease environment, which may include latent inflammatory activity and underlying axonal loss, vascular insufficiency, local gliosis, and chronic inflammatory cells, may all conspire to make adult targets less approachable than their pediatric counterparts (Franklin 2002). Indeed, as noted previously, in some disease settings endogenous progenitors may be intrinsically competent to remyelinate adult demyelinated brain, but may be impeded from doing so in aged animals by a deficient innate immune system, which might in turn be rescued by immune reconstitution from a younger parabiotic partner (Ruckh et al. 2012). In that regard, human microglia may now be produced from PSCs as well, and can engraft host brains efficiently (Abud et al. 2017; Shibuya et al. 2022), raising the possibility of co-engrafting young human OPCs and microglia together, so as to provide a more permissive immune environment for OPC engraftment and remyelination.

Clearly then, any cell-based therapeutic strategies for adult demyelination, especially those intended to remyelinate both acute and chronic lesions of MS, will require aggressive disease modification, as well as rigorous stratification to define those patients with sufficient axonal integrity as to potentially benefit from this approach. That said, these data offer hope that OPC delivery may soon serve as a viable treatment option for the restoration of both lost myelin and compromised function to afflicted patients.

SYNOPSIS

Our understanding of the biology of oligodendrocytes and their progenitors, as well as of myelin and its disorders, has increased tremendously over the past several years, aided by new technologies in genomics and stem cell biology, transgenic and chimeric animal models of disease, and imaging and cell signaling analysis. These new insights provide real hope that triggering remyelination from pharmacologically mobilized endogenous progenitors as well as using tissue- and stem cell–derived OPCs as transplantable agents for myelin replacement are each viable strategies for myelin repair, which may each soon be poised for clinical translation. We must remain cautious though, as the associated risks of these new approaches are also only now becoming appreciated. That said, given the rapidly accelerating growth in remyelination biology over the past decade, progress over the next decade seems assured to be as scientifically exciting as it promises to be therapeutically meaningful.

COMPETING INTEREST STATEMENT

R.J.M.F is an employee of Altos Labs. S.A.G. is a part-time employee and stockholder of Sana Biotechnology, from which his laboratory receives sponsored research support. S.A.G. is also a co-founder of CNS2, Inc., in which he holds equity and from which his laboratory receives research support.