The Landscape of Targets and Lead Molecules for Remyelination

Abstract

Remyelination, or the restoration of myelin sheaths around axons in the central nervous system (CNS), is a multi-stage repair process that remains a major need for millions of patients with multiple sclerosis (MS) and other diseases of myelin. Even into adulthood, rodents and humans can generate new myelin-producing oligodendrocytes, leading to the therapeutic hypothesis that enhancing remyelination could lessen disease burden in MS. Multiple labs have used phenotypic screening to identify dozens of drugs that enhance oligodendrocyte formation, and several hit molecules have now advanced to clinical evaluation. Target identification studies have revealed a large majority of these hits share the ability to inhibit a narrow range of cholesterol pathway enzymes and thereby induce cellular accumulation of specific sterol precursors to cholesterol. This Perspective surveys the recent fruitful intersection of chemical biology and remyelination and suggests multiple approaches toward new targets and lead molecules to promote remyelination.

Introduction

‘Regenerative medicine’ approaches aim to repair and reverse a wide range of disease states. In the CNS, one such regenerative approach entails remyelination, or the restoration of insulating myelin sheaths around axons. Intractable myelin diseases including MS affect millions, making remyelination a major unmet medical need. While many immunomodulatory drugs are FDA-approved to mitigate immune-mediated demyelination in MS, these agents are unable to reverse existing damage or halt disease progression. As data supporting the therapeutic promise of remyelination have strengthened over the past decade, groups across academia and industry have begun work toward the first ‘remyelinating therapeutics,’ and several candidate molecules have now advanced to clinical investigation.

This Perspective first distills the evidence supporting the therapeutic hypothesis of remyelination and then surveys approaches used in the field to identify novel targets and lead molecules for promoting remyelination. In particular, phenotypic ‘drug repurposing’ screens have identified dozens of small molecules that promote formation of oligodendrocytes from oligodendrocyte progenitor cells (OPCs), an abundant CNS stem cell population that readily differentiates to oligodendrocytes in vivo and in vitro. Numerous fundamentally similar screens across multiple labs have validated dozens of FDA-approved drugs and tool molecules that enhance formation of oligodendrocytes; these screens have also catalyzed multiple drug repurposing clinical trials in MS. Recent work indicating that nearly all hits obtained by independent labs share an ‘off-target’ ability to inhibit specific enzymes in cholesterol biosynthesis represents a unique case study in small-molecule target identification that has spotlighted altered sterol signaling as a new approach to enhancing oligodendrocyte formation. Finally, we survey recent advances in the biology of remyelination that suggest new phenotypic screening approaches to go beyond sterol pathway modulators and uncover the next generation of targets and lead molecules for promoting remyelination.

The therapeutic promise of remyelination

By wrapping (or ‘ensheathing’) its own plasma membranes around neural cell axons, oligodendrocytes accelerate neural transmission and nourish axons with metabolic support essential for long-term integrity. Underscoring myelin’s vital importance for proper CNS function, MS and other disorders of myelin cause disabling impairments to movement and cognition. Critically, remyelination has been shown experimentally, and increasingly clinically, to improve outcomes by restoring neuronal transmission and promoting survival of remyelinated axons (as reviewed previously^1–3). As such, the restoration of myelin around denuded axons is widely regarded as a promising regenerative and neuroprotective strategy in various diseases.

Among indications that could benefit from pro-remyelination therapies, most prevalent is MS (ca. 3 million patients worldwide), so called for the ‘many scars’ caused by wayward immune cell attacks on myelin. The frequent resolution of neurological symptoms during the relapsing-remitting disease stage is thought to reflect, at least in part, spontaneous remyelination; progressive disability without remission, by contrast, is theorized to reflect exhausted remyelination. Similar yet distinct neuroinflammatory demyelinating diseases include neuromyelitis optica spectrum disorders (Devic’s disease), acute disseminated encephalomyelitis, and myelin-associated glycoprotein antibody-associated disease.⁴ Notably, certain immune cells and molecules can inhibit remyelination,^5–7 which may dampen the impact of remyelinating therapies and require that patients maintain existing immunomodulatory treatments to suppress immune attacks.

Pro-remyelination therapies may also benefit diseases not primarily inflammatory in nature. In ALS, mouse genetic evidence of an unexpected role for oligodendrocytes in disease onset was corroborated in well-powered genetic studies in humans with ALS;^8,9 a strikingly similar story exists for schizophrenia.¹⁰ Emerging roles for oligodendrocytes in Alzheimer’s Disease and Parkinson’s Disease hints at broad potential for remyelinating therapeutics across a spectrum of neurodegenerative disease.^11,12 Rare but devastating genetic diseases characterized by loss of myelin, including ‘leukodystrophies’ like Pelizaeus-Merzbacher disease and vanishing white matter disease, as well as more common genetic conditions like Down Syndrome, could also benefit from remyelinating therapeutics.^13,14 Remyelination is also needed during axon regeneration after nerve injury¹⁵ and following oligodendrocyte loss due to ischemic injury, as in stroke¹⁶ or periventricular leukomalacia.¹⁷ Remyelinating therapeutics thus hold broad potential across a neuropathological spectrum.

The functional significance of remyelination

Our understanding of remyelination has emerged primarily from rodent models in which controlled demyelination follows toxin-, genetic-, or immune-mediated insults. Injection of the lipid-disrupting lysophophatidylcholine (LPC), for example, into heavily myelinated ‘white matter tracts’ induces focal demyelination of a small region of the brain or spinal cord, whereas ingestion of a copper-chelating small-molecule, cuprizone, induces widespread CNS demyelination (Fig. 1).^18,19,20,21 Likewise, extensive demyelination followed by function-restoring remyelination occurs in cats fed an irradiated diet during gestation.²² Multiple genetic approaches to selective ablation of oligodendrocytes have also been reported, including cell-specific expression of either Diphtheria Toxin or inducible caspase-9.^23,24 In contrast to toxin-induced or genetic models, experimental autoimmune encephalomyelitis (EAE) is an immune-mediated demyelinating disease induced by peripheral injection of myelin peptides paired with immune stimulation.²⁵

Figure 1: The tissue response to myelin injury requires the coordination of various cells and signals.

(a) Healthy oligodendrocytes (OLs; light purple) identified by major myelin proteins MBP, PLP, and MOG extend processes that insulate axons (gray). Myelination enhances nerve activity, depicted by higher amplitude nerve impulses. (b) Myelin injury retards neuronal signaling and initiates phagocytic clearance of myelin debris by microglia. (c) OPCs, identified by A2B5, PDGFRα, and/or NG2, respond to distinct molecular cues that orchestrate migration, proliferation, and ultimately differentiation toward oligodendrocytes. (d) Pre-myelinating oligodendrocytes engage denuded axons. (e) Remyelination by newly-formed OLs and restoration of accelerated nerve activity completes the tissue response to myelin injury.

The regenerative response to demyelination is similar no matter its proximal cause: clearance of myelin debris by microglia is necessarily the first step, since debris is a potent inhibitor of the regenerative process (Fig. 1b).^26,27 In parallel, oligodendrocyte progenitor cells (OPCs) migrate into the demyelinated area and proliferate (Fig. 1c).^28,29 These bipolar OPCs then differentiate into highly processed pre-myelinating oligodendrocytes (Fig. 1d). Once in contact with a bare axon, a distinct set of molecular cues orchestrates the vast cytoskeletal rearrangement necessary to re-invest the axon with multiple layers of lipid-rich myelin membranes (Fig. 1e).³⁰ Although the process of remyelination largely mirrors developmental myelination,^31,32 mechanisms unique to remyelination, both cell-intrinsic^33,34 and cell-extrinsic,³⁵ influence its rate and extent.

Experimental evidence strongly implies that remyelination imparts functional benefits. Following experimental demyelination of the optic nerve, recovery in visual conductance tightly correlated with remyelination,³⁶ directly linking the degradation and recovery of myelin with loss and gain of visual function. In a mouse model of optic nerve crush, the regeneration of axons lacking myelin was insufficient to restore visual function; remyelination was required.¹⁵ Indeed, visual conductance was a primary endpoint for remyelination in a recent successful MS clinical trial.³⁷ Accelerated remyelination also correlated with improved functional recovery across a swath of animal models, including immune-mediated,³⁸ genetic,³⁹ and traumatic.⁴⁰

While these functional gains likely reflect restored neural conductance, remyelination also is neuroprotective. By clustering membrane proteins (e.g., ion channels), myelin balances energy demand and supply. As such, the loss of myelin causes metabolic stress in neurons that promotes their degeneration over time;⁴¹ accelerated remyelination prevented axonal loss in vivo.⁴² Myelinating oligodendrocytes also nourish axons with metabolic building blocks such as lactate and pyruvate that power neuronal signalling. Together, these data suggest that myelin’s neuroprotective effects may also contribute to improved functional recovery.

Clinical data also support the functional significance of remyelination. Myelin-specific probes for use in positron emission tomography (PET) imaging provide the means to track myelin changes in longitudinal studies of human subjects. In a study of individuals in the relapsing-remitting stage of MS, myelin PET signal strongly correlated both with the expanded disability status score (EDSS)—a gold-standard readout of neurologic function—and the extent of thalamic atrophy, one of the few biological correlates of MS progression.⁴³ Conversely, the extent of demyelination in MS patients was associated with axon loss, providing proof-of-concept in human that myelin is essential for long-term axonal health.⁴⁴ Additionally, analysis of post-mortem tissue has provided evidence that a failure of OPC differentiation often limits remyelination. Supporting this idea, oligodendrocyte progenitor cells (OPCs) were present at the edges of MS lesions but appeared stalled in their differentiation to myelinating oligodendrocytes.^45,46 This finding suggested that endogenous remyelination is often insufficient to prevent disease progression over time and that enhancing this repair capacity by pharmacologically accelerating OPC differentiation to oligodendrocytes could represent a promising therapeutic strategy.

While OPCs are generally believed to be the direct precursor responsible for formation of new myelin, alternative routes to remyelination have recently been described. Surviving oligodendrocytes within a demyelinated region may contribute to remyelination by extending new processes to re-wrap denuded axons (a single oligodendrocyte can wrap 50 nearby axons).^47,48 Additionally, neural stem cells (NSCs) in the subventricular zone may form oligodendrocytes via an alternative differentiation pathway,⁴⁹ and Schwann cells, which canonically provide myelin in the peripheral nervous system (PNS), can arise from CNS-localized OPCs.⁵⁰ By whatever means it occurs, remyelination’s ability to restore neuronal transmission and promote axonal survival makes it a promising therapeutic strategy for MS and other disorders characterized by loss of myelin.

Target-driven approaches to identifying lead molecules

Efforts to identify enhancers of remyelination have predominantly sought to promote the differentiation of cultured OPCs to new oligodendrocytes, an approach supported by the stalled differentiation of OPCs within MS lesions.⁴⁵ Initial efforts to promote remyelination targeted existing positive and negative regulators of oligodendrocyte formation. For example, thyroid hormone (TH) has long been established as a potent driver of OPC differentiation via the thyroid hormone receptor. While physiological levels of TH are tightly regulated, a Phase 1 trial of TH confirmed its short-term safety and tolerability⁵¹ and thyromimetics including sobetirome are currently in development.⁵² The Retinoid X receptor gamma (RXRγ), a thyroid hormone receptor heterodimerization partner, was identified as another positive regulator in transcriptomic studies of rodent remyelination.⁵³ While the RXR agonist 9-cis-retinoic acid improved remyelination in rodents, a Phase 2 clinical trial of the RXRγ agonist Bexarotene missed its primary imaging endpoint but showed improved optic nerve conductance rates.⁵⁴ Progesterone receptor agonists have also been shown to promote remyelination in vivo and have advanced for preclinical development.⁵⁵

A parallel approach to accelerating remyelination entails antagonizing established negative regulators of OPC differentiation. Mouse genetic studies initially supported Lingo-1 as a negative regulator, and an anti-Lingo-1 antibody accelerated remyelination and supported axonal integrity in the EAE model of inflammatory demyelination.⁵⁶ However, clinical development of a monoclonal antibody targeting Lingo-1 (opicinumab; BIIB-033) has been stymied by lack of promising clinical efficacy. GPR17 is expressed highly in developing and injured OPCs, and genetic studies have indicated that its inactivation promotes remyelination.⁵⁷ These observations have led to characterization of a number of potential GPR17 antagonists as enhancers of oligodendrocyte formation.⁵⁸ Separately, the Gli1 family of transcription factors have also been observed to negatively regulate the ability of NSCs to promote remyelination, and Gli antagonists including GANT-61 enhance remyelination in vivo.⁴⁹ However fruitful, the relatively small number of available targets have led multiple groups to use phenotypic screening as a complementary approach to identify new lead molecules and targets for promoting remyelination.

Unbiased small-molecule screening approaches

Over the past decade, a growing array of OPC sources, genotypes, readouts, and screening technologies have been employed to identify new molecules that enhance the formation of oligodendrocytes from OPCs. Initial efforts leveraged immortalized rodent OPC-like cells, including Oli-Neu⁵⁹ and CG-4, that can be expanded extensively to support HTS but whose differentiation to mature oligodendrocytes is compromised and whose use is now uncommon.⁶⁰ Later studies leveraged primary rodent OPCs obtained from dissociated mouse or rat brain that retain the ability to differentiate to mature oligodendrocytes over 3–5 days in vitro.^61–63 However, these cells are challenging to obtain in abundance for screening due to a finite rodent supply and limited expansion capacity. Pluripotent stem cell (PSC)-derived OPCs provide a scalable, pure source of OPCs that readily differentiate to mature oligodendrocytes and also enable genotypic flexibility, which has been leveraged to screen OPCs containing a leukodystrophy-inducing mutation.^64,65 The differentiation of human OPCs can also be assessed, although formation of mature oligodendrocytes generally requires 2–4 weeks, limiting feasibility for screening applications.

The differentiation of OPCs to oligodendrocytes proceeds via numerous intermediate cell states defined by the expression of lipid and protein markers (Fig. 1), several of which have been used as primary endpoints in small-molecule screens. Assays using immortalized cells typically relied on immunostaining for O4, a lipid expressed on the surface of lineage-committed but immature oligodendrocytes.^59,60 In contrast, studies using primary and PSC-derived OPCs have predominantly used myelin basic protein (MBP) positivity as a canonical readout of mature oligodendrocytes.^{61–63,66–68} Beyond molecular markers, studies have also explored different screening technologies, including compressed silica ‘micropillars’ that represent a unique approach to monitoring both oligodendrocyte differentiation and ‘myelination’ in 96-well format.⁶²

Drug repurposing screens identified overlapping hits

Using the approaches described above, numerous groups reported similar high-throughput screens that identified small molecules that enhance the formation of oligodendrocytes from cultured OPCs.^{61–63,66–68} These screens relied on heavily-overlapping libraries of FDA-approved drugs and related bioactive molecules, quantitated MBP+ positivity, and generally used thyroid hormone (T3) as positive control. Although these screens used differing OPC sources and screening methodologies, a reassuringly high concordance among hits was observed (Table 1). Common observations included antimuscarinics, selective estrogen receptor modulators, antifungals, and antipsychotics as enhancers of oligodendrocyte formation, and specific molecules within and beyond these target classes recurred across screens (Table 1). While not all studies provide details needed to assess hit rate, multiple screens reported a hit rate of ca. 5%, suggesting that a truly remarkable fraction of existing approved drugs could potentially be repurposed for remyelination. Very recently, a pre-print reported the first use of human PSC-derived OPCs to screen 2,400 approved drugs and known bioactive small molecules.⁶⁹ Among 21 enhancers of human oligodendrocyte formation described, 7 were previously identified in screens of rodent OPCs, indicating strong concordance across rodent and human OPC screens.

Table 1.

Screening hits identified from bioactive small molecule libraries as enhancing formation of oligodendrocytes from OPCs.

Validated hit	Validation		Canonical target	Sterol pathway target
	In vitro	In vivo
Benztropine61,62,63,66,67	m/r/h OPCs	EAE, CPZ	Muscarinic receptor	EBP67
Clemastine61,62,63,67	m/r/h OPCs	EAE, LPC	Histamine/muscarinic	EBP67
Tamoxifen63,66,67	m/r/h OPCs	LPC	Estrogen receptor	EBP67,80
Bazedoxifene77	m/r OPCs	LPC	Estrogen receptor	EBP67,77,80
Miconazole66,67	m/h OPCs	EAE, LPC	Fungal CYP51	CYP5167
Ro 25–698165	m OPCs	Jimpy mouse	NMDA receptors	S14R87
Quetiapine62,63	m/r/h OPCs	LPC	Atypical antipsychotic	EBP87
U5048861,67	m/r OPCs	LPC	k-opioid receptor	EBP67
Thyroid Hormone (T3)	m/r/h OPCs	EAE, CPZ	Thyroid hormone receptor	None
Clobetasol66	m OPCs	EAE, LPC	Glucocorticoid receptor	None
GSK23951278	m/r OPCs	CPZ	Histamine H3 receptors	EBP87
Vitamin C68	m NPCs	CPZ	Dioxygenases	None
Hydroxyzine66,67	m/r OPCs		Histamine receptors	EBP67,80
Raloxifene63,67	m/r/h OPCs		Estrogen receptor	EBP67,80
Toremifene63,67	m/r OPCs		Estrogen receptor	EBP67,80
Ketoconazole61,66,67	m/r/h OPCs		Fungal CYP51	CYP5167
Medroxyprogesterone acetate66,67	m OPCs		Progesterone receptor	CYP5167
Megestrol66,67	m OPCs		Progesterone receptor	CYP5167
Ifenprodil63,65,66,67	m/r OPCs		NMDA receptors	S14R67
Fluphenazine61,63	m/r/h OPCs		Antipsychotic; multiple receptors	EBP67,80
Salmeterol61,63	r OPCs		Adrenergic receptors	S14R87
Vanoxerine61,63	m/r OPCs		Multiple ion channel families	EBP87
Donepezil63,66	m/r OPCs		Acetylcholinesterase	EBP87
Vesamicol63,66,67	m/r OPCs		Vesicular ACh transport	EBP67
SB36679183	m OPCs		TRP channel antagonist	HSD17B783
All-trans-retinoic acid61	r OPCs		Retinoic acid receptor	None
Parbendazole (p57Kip2)73	m/r/h OPCs		Tubulin assembly	None
Danazol (p57Kip2)73	m/r/h OPCs		Estrogen receptor	None

While these hits span a diverse range of established cellular targets, many share the ability to inhibit specific enzymes in cholesterol biosynthesis. Unless noted, these studies used MBP as a marker of mature oligodendrocytes. “m/h/r” refers to mouse, rat, or human OPCs. LPC, lysophosphatidyl choline; CPZ, cuprizone; EAE, experimental autoimmune encephalitis.

Forging a standard screening cascade

Prioritizing the large number of validated hits emerging from these screens has necessitated numerous secondary and tertiary assays, and a standard ‘screening cascade’ has emerged (Fig. 2). Following confirmation of enhanced oligodendrocyte formation in the primary assay, several groups have demonstrated small-molecule-mediated enhancement of myelination using co-culture of OPCs with neurons derived from spinal cord dorsal root ganglia⁷⁰ or inorganic ‘microfiber’ plates as axon-like surrogates on which to culture OPCs.⁷¹ Evaluation of human OPC differentiation is another common secondary assay.^72,73 Additionally, human oligodendrocyte and myelin formation have been assessed in oligocortical spheroids, a ‘brain organoid’ system modified to favor the development and differentiation of OPCs within human tissue.⁷⁴ Finally, many leading hits have been further characterized as enhancing remyelination in vivo using the LPC, cuprizone, and EAE models described above. Several screening hits characterized in these assays have now advanced for clinical evaluation in MS patients, including clemastine, which induced statistically significant improvement in optic nerve conductance rates,³⁷ and bazedoxifene (ClinicalTrials.gov Identifier: NCT04002934). Ultimately, these parallel screens of rodent OPC differentiation using FDA-approved drug libraries established a now widely-used screening cascade that has validated dozens of enhancers of oligodendrocyte formation and remyelination.

Figure 2. A general assay flowchart for identifying small-molecule enhancers of remyelination.

While reported screens have used varying approaches to identify and prioritize lead molecules, these assays remain widely-used choices for further characterizing his emerging from screens of OPC differentiation.

Target ID: many cellular targets

While resolving the cellular targets of phenotypic screening hits is typically challenging, screening libraries of FDA-approved drugs can simplify target elucidation since most drugs have established cellular targets. Several early publications provided evidence that individual hits functioned in OPC by modulating their canonical drug target. Muscarinic receptor antagonists were prominent hits in several early screens, and subsequent work used mouse genetics to demonstrate that loss of the M1 muscarinic receptor but not other subtypes was associated with enhanced oligodendrocyte formation.³⁸ However, many other antimuscarinics did not affect oligodendrocyte formation.⁶¹ A similar mouse genetics approach was subsequently used to implicate the kappa-opioid receptor as the functional target by which U-50488 acts in OPCs.⁷⁵ On the basis of siRNA studies, the selective estrogen receptor modulator (SERM) tamoxifen was suggested to function via the estrogen receptor alpha, estrogen receptor beta, and GPR30.⁷⁶ However, later work suggested that the SERM bazedoxifine functioned independent of estrogen receptors.⁷⁷ Chemical-genetic and genetic evidence was also used to support the histamine H3 receptor as the functional target of GSK293512 and related histamine H3 modulators that enhance oligodendrocyte formation.⁷⁸

A unifying cellular activity in OPCs

While initial studies implicated the canonical targets of these drugs as responsible for enhanced oligodendrocyte formation, available data in several cases left room for alternate interpretations. First, the sheer number of well-validated hits spread across highly diverse target classes raised the question of whether dozens of established drug targets could all equally regulate the OPC differentiation process (Table 1). Additionally, numerous targets implicated as enhancing OPC differentiation were not clearly expressed in publicly available RNAseq and single-cell RNAseq data obtained from OPCs.^66,79

These observations led to the distinct hypothesis that many of these hits could share an ‘off-target’ effect or signal through a unifying downstream signaling node. Early observations consistent with this idea included that well-validated enhancers of oligodendrocyte formation shared a maximal effect size and that pairwise combinations of these validated hits showed no additive effects, as might be expected if the molecules were working through independent cellular mechanisms.⁶⁷

8,9-unsaturated sterols drive oligodendrocyte formation

Recent work has established that more than 5% of FDA-approved drugs inhibit cholesterol biosynthesis.⁸⁰ Strikingly, many well-validated enhancers of oligodendrocyte formation were already established to inhibit cholesterol biosynthesis (Table 1), particularly the enzymes CYP51 and EBP (Figure 3). Multiple lines of evidence now support that inhibition of specific cholesterol biosynthesis enzymes promotes oligodendrocyte formation and that this activity accounts for a large majority of validated hits identified by various labs. Initial chemical-genetic studies using selective inhibitors of 8 cholesterol pathway enzymes established that inhibition of many pathway enzymes was insufficient to promote oligodendrocyte formation; rather, only a narrow window of enzymes in the center of the pathway spanning CYP51 to EBP promoted oligodendrocyte formation upon blockade (Figure 3).⁶⁷ Structurally unrelated series of inhibitors of CYP51, sterol 14-reductase, and EBP were equally effective at promoting oligodendrocyte formation, and potency for enhancing OPC differentiation closely matched potency for target inhibition in OPCs. (Well-established GCMS and LCMS techniques enable quantitation of nearly all cholesterol precursors in parallel, providing a rapid ‘sterolomics’ approach to gauge both target inhibition and target selectivity of small molecule and genetic perturbations in any cell type or tissue.⁸¹) Additionally, targeting CYP51 using siRNA and EBP using CRISPR/Cas9 enhanced oligodendrocyte formation. As described further below, dozens of validated hits from multiple screens shared the ability to inhibit these pathway enzymes in OPCs, lending further support (Table 1). Finally, dosing regimens established to promote remyelination in mouse models of MS also elevated levels of the 8,9-unsaturated sterol substrates of these enzymes in mouse brain, indicating that these efficacious treatments inhibited cholesterol synthesis within the CNS.⁶⁷

Figure 3: The cholesterol biosynthesis pathway.

Cholesterol biosynthesis begins with acetyl CoA, proceeds through the parallel Kandutsch-Russell, Bloch, and epoxycholesterol shunt pathways.^67,100 Lanosterol and all sterols within the gray box contain the 8,9-unsaturation associated with enhanced oligodendrocyte formation. Enzymes whose inhibition induces 8,9-unsaturated sterol accumulation and oligodendrocyte formation are orange; cholesterol precursors shown to enhance oligodendrocyte formation are labelled in blue.

Inhibition of enzymes spanning CYP51 to EBP is unique in inducing cellular accumulation of cholesterol precursors that share a specific structural feature, the Δ8,9 olefin (Fig. 3). These 8,9-unsaturated sterols are typically maintained at levels one-one hundredth those of cholesterol in cells and tissues; however, inhibition of an appropriate pathway enzyme can lead to rapid and robust accumulation of its 8,9-unsaturated substrates due to the linear nature of cholesterol biosynthesis (Fig. 3). 8,9-unsaturated sterols were previously shown to promote oocyte maturation,⁸² suggesting that these cholesterol precursors could also serve as signaling molecules to promote conversion of OPCs to oligodendrocytes. In support of this idea, nine purified 8,9-unsaturated sterols were sufficient to drive oligodendrocyte formation from OPCs. Conversely, preventing cellular accumulation of these sterols abrogated a pro-differentiation effect, supporting accumulation of 8,9-unsaturated sterols as the direct mechanism by which inhibition of these enzymes enhanced oligodendrocyte formation.⁶⁷ The downstream signaling effects of 8,9-unsaturated sterol accumulation remain unclear but could provide important understanding of effector pathways promoting oligodendrocyte maturation and suggest new targets downstream of the cholesterol pathway.

Subsequent work expanded the range of pathway targets and sterols that promote oligodendrocyte formation. Two additional enzymes with 8,9-unsaturated sterol substrates, SC4MOL and HSD17B7, also promoted oligodendrocyte formation when targeted via CRISPR or with newly-developed potent small-molecule inhibitors.⁸³ Additional studies used genetic and chemical-genetic approaches to establish that partial inhibition of lanosterol synthase, LSS, enhanced oligodendrocyte formation by promoting flux through the ‘epoxycholesterol shunt’ and inducing cellular accumulation of epoxycholesterol.⁸⁴ As epoxycholesterol lacks an 8,9-unsaturation, this finding establishes that some sterols beyond the 8,9-unsaturated class can also promote OPC differentiation to oligodendrocytes. However, recent work screened a focused library of bioactive sterols and steroids but did not find additional effective classes, suggesting that only a specific, limited range of sterols retain the ability to influence oligodendrocyte formation.⁸⁵

A dominant mechanism for dozens of HTS hits

Remarkably, a large majority of enhancers of MBP+ oligodendrocyte formation share the ability to inhibit this set of cholesterol pathway enzymes. An initial report characterized 15 validated hits obtained by multiple labs, including clemastine, benztropine, and miconazole, as uniformly inhibiting CYP51, sterol 14 reductase, or EBP, with EBP most frequently targeted.⁶⁷ Subsequent work supported that nearly all validated hits from a wide range of screens induced accumulation of 8,9-unsaturated sterols regardless of the screening library or source of OPCs used, including human OPCs.^86–88 These analyses suggest that 90+% of validated hits from screens for enhancers of MBP+ oligodendrocyte formation from rodent and human OPCs target cholesterol biosynthesis at enzymes spanning CYP51 and EBP. To explain why such a remarkable fraction of FDA-approved drugs have affinity for cholesterol biosynthesis enzymes like EBP,⁸⁷ recent work noted both the high inherent druggability of cholesterol pathway enzymes, which contain large hydrophobic substrate-binding pockets, and the ability of many tertiary amine-containing molecules to mimic a sterol cation-like transition state in EBP’s enzymatic mechanism.

While many screening hits share the ability to induce 8,9-unsaturated sterol accumulation, other validated hits have no impact on cholesterol synthesis. As noted earlier, thyroid hormone is a widely-used and highly effective positive control in assays monitoring OPC differentiation and has no effect on cellular sterol levels. Other nuclear receptor ligands, including retinoids and glucocorticoids, have also been obtained by multiple groups and shown to have no impact on cholesterol biosynthesis (Table 1). Interestingly, a recent screen that sought to enhance formation of early oligodendrocytes from OPCs identified a distinct spectrum of hits with no known links to cholesterol biosynthesis, suggesting that enrichment for sterol pathway modulators may in part relate to the assay endpoint used (i.e., mature MBP+ oligodendrocytes).⁷² Use of distinct endpoints, screening libraries, and assay conditions are also likely to capture unique hit classes beyond sterol pathway modulators (see ‘Identifying Next-gen pro-myelinating targets and leads’, below).

Sterols as influencers of oligodendrocyte biology

The ability of 8,9-unsaturated sterols to enhance differentiation of OPCs is one of several ways that sterols influence the course of myelin formation. Myelin membranes are highly enriched in cholesterol, and the blood-brain-barrier requires that this cholesterol is synthesized by oligodendrocytes or trafficked to them by neighboring cells. The ability of oligodendrocytes to synthesize sterols is essential for myelin formation,⁸⁹ as genetic inactivation of squalene synthase (SQS/FDFT1), an upstream pathway enzyme needed for synthesis of all sterols (Fig. 3), impairs myelination. Additional in vivo studies have recently shown that ‘recycling’ of cholesterol from myelin debris influences remyelination.⁹⁰ Following phagocytosis of myelin debris by macrophages and microglia, cholesterol can be redistributed via lipoprotein particles to support the needs of remyelinating oligodendrocytes. Myelin debris-laden phagocytes also trigger signaling pathways including LXR that promote cholesterol efflux and limit inflammation within the demyelinated lesion. Interestingly, this recycling process may become inefficient in the context of chronic demyelinated lesions, forcing oligodendrocytes to rely on de novo synthesis and ultimately contributing to remyelination failure.⁹¹

Given the essential role of sterol synthesis during myelin formation in vivo, the ability of 8,9-unsaturated sterols to signal to promote oligodendrocyte formation is somewhat surprising. However, multiple observations suggest that perturbing cholesterol synthesis with small-molecule inhibitors of EBP or CYP51 may have effects distinct from genetic inactivation of SQS. First, in contrast to genetic manipulations, an inhibitor would partially and temporarily inhibit its target to induce accumulation of 8,9-unsaturated sterols while also enabling some cholesterol synthesis. Additionally, because SQS lies far upstream in the cholesterol pathway, loss of this enzyme prevents synthesis of all cellular sterols. In contrast, small-molecule inhibition of EBP maintains total cellular sterol levels but favors a ‘sterol shift’, with accumulation of the EBP substrate zymostenol counterbalanced by reduction of cholesterol. Whether sterol precursors like zymostenol—which differs from cholesterol only in the placement of a single double bond—can be incorporated into functional myelin is unknown, though other sterol precursors to cholesterol have previously been found to accumulate in myelin.⁹² Finally, numerous studies have demonstrated that molecules that induce accumulation of 8,9-unsaturated sterols promote recovery from paralysis in the MOG-EAE model, consistent with functional remyelination in vivo. While these observations support the ability of certain cholesterol pathway modulators to enhance functional remyelination, future work will be required to evaluate how this altered sterol environment influences various stages of the remyelination process.

Identifying next-gen pro-myelinating targets and leads

Most studies to date have sought to promote remyelination by identifying cell-intrinsic enhancers of oligodendrocyte formation from OPCs, and these studies have predominantly identified inhibitors of cholesterol synthesis that induce 8,9-unsaturated sterol accumulation. However, encouraging OPCs to mature to oligodendrocytes represents only a single approach toward modulating the complex, multi-stage biology of remyelination, and even within this approach additional targets for promoting oligodendrocyte formation doubtless remain to be identified. Here we suggest future phenotypic screening strategies to identify additional targets and lead molecules for promoting remyelination.

First, modifications to the well-established assay measuring differentiation of OPCs to oligodendrocytes could be employed to bias hits away from cholesterol pathway enzymes and toward novel activities. Future screens could rapidly triage hits modulating the sterol pathway using gas chromatography / mass spectrometry, or screens could be run in the presence of a saturating concentration of an EBP inhibitor. Alternatively, genetic rather than chemical-genetic screens could be employed to identify novel targets in an unbiased fashion. The limited scale afforded by primary OPCs could potentially enable screens of focused subsets of the genome (kinome, epigenome, etc.), while the enhanced expandability of pluripotent stem-cell-derived OPCs may open the door to genome-wide screening approaches.

Second, OPC differentiation assay conditions could be modified to better reflect a disease context. The demyelinated lesion microenvironment contains many factors that inhibit OPC differentiation and remyelination, including various pro-inflammatory molecules, extracellular matrix proteins,^35,93 and myelin debris.²⁷ Age-related factors also impede remyelination, including decreased metabolism and increased stiffness of the extracellular niche.^34,94,95 Several validated enhancers of remyelination were ineffective in the context of aged OPCs, suggesting that future screens performed in the presence of aged OPCs and/or other inhibitory factors have the potential to identify a unique spectrum of hits.^95,96

Small molecules that accelerate distinct phases of myelinating oligodendrocyte formation may also reveal distinct targets and patterns of efficacy. While most studies to date have focused on promoting formation of mature MBP+ oligodendrocytes, a recent screen sought to accelerate the initial differentiation of OPCs to early oligodendrocytes (marked by cytoplasmic p57kip2) to capture distinct hit classes.⁷³ Future work may expand on this approach or alternatively seek to modulate later stages of maturation, including the conversion of premyelinating oligodendrocytes to myelinating oligodendrocytes. While various low-throughput myelination assays are well-established (Fig. 2), new technologies may enable higher-throughput evaluation of ‘myelination’ in culture. In particular, electrospun microfibers that can be ensheathed by cultured oligodendrocytes are available in microplate format,⁷¹ and recent work has also used 3D-printing to develop synthetic substrates for assaying ‘myelination’ in cell culture.⁹⁷ Future screens could use these specialized microplates to evaluate myelin thickness (the “g-ratio”), the physical length of the myelin (the “internode”), the number of fibers wrapped per oligodendrocyte, or other parameters.

While each approach above seeks to directly target OPCs or oligodendrocytes, modulating cell types beyond the oligodendrocyte lineage may provide an indirect route to remyelination. Given that debris clearance is rate-limiting for remyelination and that small molecule promoters of myelin phagocytosis enhance remyelination,⁹⁶ screens that more broadly identify small molecules that enhance the phagocytic function of microglia/macrophages could be valuable. Additionally, recent work has demonstrated that reactive astrocytes can induce death of mature oligodendrocytes and are common in MS and other neurodegenerative diseases.⁹⁸ Identifying small molecules that modulate reactive astrocyte states could indirectly favor remyelination by promoting survival of newly formed oligodendrocytes and suppressing release of inflammatory cytokines known to limit remyelination.

Finally, in vivo screening approaches have also been used to evaluate myelination, most extensively using optically transparent zebrafish. Impressively, screening an epigenetics-focused library of 175 small molecules yielded four enhancers of oligodendrocyte formation in zebrafish, and the automated screening method suggests larger screens may be feasible.⁹⁹ While this approach potentially enables identification of modulators of any stage of the remyelination process within the highly relevant in vivo context, the ability of hits emerging from these species to promote remyelination in rodents remains to be broadly assessed.

Conclusion

Over the past decade, mounting evidence supporting the feasibility and functional relevance of remyelination to MS and other myelin disorders has inspired the field to identify new therapeutic targets and lead molecules to advance drug discovery efforts. Parallel efforts by many labs to use phenotypic drug repurposing screens to identify enhancers of differentiation of OPCs to mature oligodendrocytes have forged a typical screening cascade, and validated hits have improved remyelination in rodent models and advanced to clinical evaluation. The convergence of many such hits on specific cholesterol biosynthesis enzymes has revealed a new role for sterol metabolites as enhancers of oligodendrocyte formation and suggested new targets for drug discovery. Beyond directly enhancing oligodendrocyte formation from OPCs, complementary screening approaches that mimic aspects of the demyelinated microenvironment or modulate cell types other than OPCs appear poised to identify the next wave of targets and leads for promoting remyelination. Continued rapid progress in the field suggests that first-in-class therapies poised to deliver on the therapeutic promise of remyelination are on the horizon.