Deciphering the role of sphingosine 1‐phosphate in central nervous system myelination and repair

Abstract

Sphingosine 1‐phosphate (S1P) is a bioactive lipid of the sphingolipid family and plays a pivotal role in the mammalian nervous system. Indeed, S1P is a therapeutic target for treating demyelinating diseases such as multiple sclerosis. Being part of an interconnected sphingolipid metabolic network, the amount of S1P available for signalling is equilibrated between its synthetic (sphingosine kinases 1 and 2) and degradative (sphingosine 1‐phosphate lyase) enzymes. Once produced, S1P exerts its biological roles via signalling to a family of five G protein‐coupled S1P receptors 1–5 (S1PR_1–5). Despite significant progress, the precise roles that S1P metabolism and downstream signalling play in regulating myelin formation and repair remain largely opaque and somewhat controversial. Genetic or pharmacological studies adopting various model systems identify that stimulating S1P‐S1PR signalling protects myelin‐forming oligodendrocytes after central nervous system (CNS) injury and attenuates demyelination in vivo. However, evidence to support its role in remyelination of the mammalian CNS is limited, although blocking S1P synthesis sheds light on the role of endogenous S1P in promoting CNS remyelination. This review focuses on summarising the current understanding of S1P in CNS myelin formation and repair, discussing the complexity of S1P–S1PR interaction and the underlying mechanism by which S1P biosynthesis and signalling regulates oligodendrocyte myelination in the healthy and injured mammalian CNS, raising new questions for future investigation.

Sphingosine‐1‐phosphate (S1P) signals to a family of five S1P receptors 1–5 (S1PR_1–5) coupled with different G proteins. Current studies indicate that S1P exerts differential effects on oligodendrocytes (OLG) and myelin depending on the receptor being activated: promoting the survival of oligodendrocyte precursor cells (OPC) via activating S1PR₁ whilst potentiating the survival of mature OLG via S1PR₅.

Abbreviations

CNS

central nervous system

CS

ceramide synthases

EAE

experimental autoimmune encephalomyelitis

ERK

extracellular signal‐regulated kinase

fingolimod‐P

fingolimod‐phosphate

MAPK

mitogen‐activated protein kinases

MS

multiple sclerosis

OLG

oligodendrocytes

OPCs

oligodendrocyte precursor cells

PNS

peripheral nervous system

S1P

sphingosine 1‐phosphate

S1PR₁

S1P receptor subtype 1

S1PR₂

S1P receptor subtype 2

S1PR₃

S1P receptor subtype 3

S1PR₄

S1P receptor subtype 4

S1PR₅

S1P receptor subtype 5

SMase

sphingomyelinases

SMSyn

sphingomyelin synthase

SphK

sphingosine kinase enzyme

SPPs

S1P‐specific phosphatases

SPT

palmitoyltransferase

1. INTRODUCTION

Myelin is a multilayered lipoprotein‐rich membrane that surrounds many nerve fibres in the nervous system, enabling rapid and efficient nerve signal transduction along the axon (Kalakh & Mouihate, 2024). It is essential to normal motor, sensory and cognitive function of the mammalian nervous system. Myelin is produced by oligodendrocytes (OLG) in the central nervous system (CNS) and Schwann cells in the peripheral nervous system (PNS), aiding in the cell–cell interactions between neurons (Kister & Kister, 2022). The process of CNS demyelination, marked by damage to the myelin sheath or the myelin‐forming OLG, often occurring as a result of autoimmune mechanisms or ageing, causes a range of neurological deficits, such as difficulties with voluntary muscle coordination, sensorimotor functions, and cognition (Ghasemi et al., 2016; Song et al., 2021). CNS myelination comprises a series of cellular events, including the proliferation and migration of oligodendrocyte precursor cells (OPCs), followed by differentiation into mature OLG and subsequent axonal ensheathment by myelin membranes (Mitew et al., 2014). This program is tightly controlled by intrinsic factors in response to extracellular cues that could exert stage‐specific effects or regulate the full myelinating process (Mitew et al., 2014; Nave & Werner, 2014).

Lipids are essential for myelin formation. Approximately, 70%–85% of the myelin sheath is composed of lipids, whilst proteins account for only about 15%–30% (Poitelon et al., 2020). This unusually high lipid composition in myelin is contributed by cholesterols, phospholipids, such as phosphatidylcholine, sphingomyelin and plasmalogen, and glycolipids like galactosylceramide, with the three classes of lipids being present in the ratio 4:4:2; although, this ratio varies between the CNS and PNS (García‐García et al., 2024; Poitelon et al., 2020). However, because of the challenges associated with the measurement and manipulation of lipid metabolites, to what extent lipid signalling molecules regulate myelination remains poorly understood compared with the large volume of knowledge gained from proteins and amino acid‐based molecules (Barnes‐Vélez et al., 2023). Sphingosine‐1‐phosphate (S1P) is a product of sphingomyelin metabolism, which is known to play an important role in a plethora of biological processes (Proia & Hla, 2015). The effects of S1P are not only limited to signalling in higher vertebrates within specific systemic contexts but also are implicated in the developmental and physiological processes across various organ systems (Maharaj et al., 2020). The major pro‐proliferative nature of S1P is essential for the development of vasculature in the cardiovascular system, regulation of T‐cell trafficking in the immune system and neural development within both the PNS and CNS (Proia & Hla, 2015). As such, the altered S1P signalling can be a biomarker for a range of pathologies, with either multi‐systemic effects or more localised consequences.

In the CNS, S1P regulates brain structure and function during development and after injury (van Echten‐Deckert, 2023). S1P signalling plays a major role in regulating homeostatic mechanisms, markedly during different phases of development (Cartier & Hla, 2019; Zhao et al., 2015). The maintenance of the blood–brain barrier (BBB) and the differential effects on the survival and differentiation of neurons and glial cells elicited by S1P have been well established, whilst other intermediates of the signalling pathway are also gaining interest (Xiao, 2023). Whilst the role of S1P in neurotoxicity (Hagen et al., 2009; Hagen‐Euteneuer et al., 2012) or neuroprotection (Ceccom et al., 2014; Couttas et al., 2014) is still a subject of debate, the disruption of S1P signalling within the CNS is seen to have a causal relationship with the progression of major autoimmune diseases and neurodegenerative conditions, such as rheumatoid arthritis, multiple sclerosis (MS), Alzheimer's disease, and even Parkinson's disease (Ghasemi et al., 2016; Xiao, 2023).

The significance of S1P in the CNS is highlighted in its role in the therapeutic application of demyelinating diseases such as MS. Over the last 10 years, a series of drug therapies that target S1P or its signalling pathway have been developed for treating MS (Chun et al., 2021). For example, Fingolimod has been widely prescribed for the treatment of relapse‐remitting MS. More recently, ozanimod, ponesimod, and siponimod have gained approval after demonstrating promising effects (Pol et al., 2023; Xiao, 2023). Despite significant progress, however, the precise roles that S1P and its downstream signalling play in regulating myelination and remyelination remain largely unclear and somewhat controversial. Recent studies adopting in vitro, ex vivo, or in vivo model systems show that stimulating S1P signalling protects OLG and myelin loss after CNS injury; however, evidence to support its role in remyelination of the mammalian CNS is limited.

With the growing recognition of S1P as a key player in myelin formation, investigating its biosynthetic and signalling pathways holds immense promise for uncovering novel therapeutic strategies. Moreover, the emergence of novel therapies engineered to combat the effects of disrupted S1P mechanisms for the treatment of MS underscores the urgency and relevance of synthesising current knowledge in this area. Advancing our understanding of S1P lipid metabolism involved in myelination could aid the development of more targeted therapies for demyelination and neurodegeneration. In this review, we elucidate the accumulating knowledge concerning the metabolism of S1P and downstream signalling in myelination of the mammalian CNS and discuss the intricate mechanisms by which S1P influences this process, identifying potential avenues for therapeutic intervention towards improving treatment outcomes for demyelinating diseases such as MS.

2. BIOSYNTHESIS OF S1P IN AN INTERCONNECTED NETWORK

Extensive literature has documented the metabolic pathways for S1P biosynthesis. The S1P molecule is endogenously synthesised via two major routes, namely, de novo and salvage pathways (Figure 1a). The de novo pathway, occurring in the ER, is initiated by serine palmitoyltransferase (SPT), followed by subsequent enzyme reactions to produce ceramide (Merrill, 2011). The ceramide intermediate is then metabolised by ceramidase and converted into sphingosine, which is then phosphorylated by either of the two isoforms of the sphingosine kinase enzyme (SphK), particularly SphK1 and SphK2, resulting in the formation of S1P (Bravo et al., 2022; Fohmann et al., 2023). This process is reversible since S1P can be dephosphorylated back into sphingosine by S1P‐specific phosphatases (SPP) starting an energy‐consuming recycling/salvage pathway (Figure 1a). Following a dephosphorylation process, sphingosine can be metabolised back into ceramide primarily by ceramide synthase (CS) (van Echten‐Deckert, 2023). This salvage pathway enables ceramide to generate sphingolipids. Therefore, ceramide represents a critical intermediate molecule in the production of sphingosine and S1P. In many tissues, S1P can be irreversibly degraded by sphingosine‐1‐phosphate lyase (SPL) to form (2E)‐hexadecenal and ethanolamine phosphate (Figure 1a), thereby permanently removing S1P from the sphingolipid pool (Saba & Hla, 2004; Serra & Saba, 2010). The irreversible degradation of S1P by a single SPL enzyme leads to low S1P tissue levels and ultimately contributes to the fluid‐tissue gradient of this bioactive lipid, maintaining high S1P in the bloodstream including blood and lymph plasmas whilst a low level in tissues (Schwab et al., 2005). In addition, the bioactivity of S1P requires chaperone proteins that bind and transport S1P in circulatory and interstitial fluids, facilitating the activation of downstream S1P signalling (Cartier & Hla, 2019). S1P is exported by specific transporters, allowing diffusion of the lipid towards other cells or circulatory fluids, thereby enabling its extracellular signalling to its receptors expressed on the same cell in an autocrine manner or on different cells in a paracrine manner (Cartier & Hla, 2019).

FIGURE 1.

S1P regulates oligodendrocyte development and myelin formation. (a) A simplified schematic showing the synthetic pathways of S1P. The de novo synthesis of S1P involves the formation of ceramide, initiated by serine SPT. S1P can also be dephosphorylated back to sphingosine via a recycling/salvage pathway. The amount of S1P available for extracellular and intracellular signalling is equilibrated between its synthetic (SphK1/2) and degradative (SPL) enzymes. (b) Schematic showing the current understanding of S1P signalling in oligodendrocyte myelination. S1P signals to a family of five S1P receptors 1–5 (S1PR_1–5) coupled with different G proteins. Current studies indicate that S1P exerts differential effects on oligodendrocytes and myelin depending on the receptor being activated: Promoting the survival of oligodendrocyte precursor cells via activating S1PR₁ whilst potentiating the survival of mature oligodendrocytes via S1PR₅. CS, ceramide synthases; OLG, oligodendrocytes; OPCs, oligodendrocyte precursor cells; S1P, sphingosine 1‐phosphate; S1PRs, S1P receptors; SMase, sphingomyelinases; SMSyn, sphingomyelin synthase; SphK, sphingosine kinase enzyme; SPPs, S1P‐specific phosphatases; SPT, palmitoyltransferase. The figure is prepared using BioRender with a publication licence KQ270850IU.

3. S1P SIGNALLING: COMPLEXITY AND PLASTICITY

3.1. Ligand and receptor interactions

Intracellular S1P regulates targeted proteins that influence various conserved functions including mitochondrial assembly and function (Alvarez et al., 2010), chromatin structure, and transcription (Hait et al., 2009). Once released to the extracellular space by specific transporters, S1P signals through five high‐affinity G protein‐coupled receptors, S1PR₁–S1PR₅ (Grassi et al., 2019) (Figure 1b), variably expressed in many organs and cells (Bryan & Del Poeta, 2018; Kays et al., 2012; Rosen et al., 2009). S1P receptors 1–3 (S1PR_1–3) are expressed by most cells and tissues. S1PR₁ is virtually expressed in all cells whilst S1PR₄ is predominant in immune cells (Gaire et al., 2018). S1PR₁ has been extensively studied. It is a master regulator of lymphocyte egress from the lymph node and an established drug target for autoimmune demyelinating diseases such as MS (Zhao et al., 2022). In the CNS, S1PR₁, S1PR₃, and S1PR₄ are present in neurons and glial cells (Alam et al., 2023; Chun et al., 2019; Skoug et al., 2022). S1PR₁ stands out as the most abundant subtype expressed in neurons compared with the other four S1PRs (Kays et al., 2012). S1PR₂ binds to S1P with high affinity (Okazaki et al., 1993) and is primarily expressed in CNS endothelial cells (Cruz‐Orengo et al., 2014). S1PR₃ is most highly expressed in astrocytes whilst S1PR₅ is expressed almost exclusively and abundantly by the myelin‐forming OLG (Jaillard et al., 2005; Yu et al., 2004; Zhang et al., 2014), although low expression is also detected in neurons (Zhang et al., 2014). In addition to the broad expression profiles and various densities in CNS cells, the S1P–S1PR signalling could also be sexually dimorphic (Cruz‐Orengo et al., 2014), depending on the specific receptor being activated. For example, the level of S1PR₂ is significantly higher in the disease‐susceptible CNS regions of female experimental autoimmune encephalomyelitis (EAE) mice and female patients with MS than their male counterparts (Cruz‐Orengo et al., 2014), indicating S1PR₂ could be a sex‐specific molecule that regulates S1P function in the context of MS. This finding is particularly important to MS since this disease is well‐known to have a higher incidence rate in women compared with men.

Although S1PRs have a highly conserved sequence, they are coupled with multiple heterotrimeric G proteins. S1PR₁ couples exclusively with Gi protein to inhibit cAMP production, whilst S1PR₂ and S1PR₃ couple with Gi/o, Gq, and G_12/13, S1PR₄ and S1PR₅ couple with Gi/o and G_12/13 (Bryan & Del Poeta, 2018; Yuan et al., 2021). S1PR₁ and S1PR₅ share a similar coupling mechanism with Gi, leading to receptor activation (Yuan et al., 2021). Upon ligand activation, S1PRs subsequently signal to various downstream intracellular pathways well documented in the literature including Ras/extracellular signal‐regulated kinases (ERK), PLC, P13K/Akt, and Rho/Rock signalling cascades (Bryan & Del Poeta, 2018; Rosen et al., 2013). Therefore, the expression profile of G proteins required for the active execution of individual S1PR receptors could influence the efficacy of current S1P analogues or S1PR modulators, many of which are considered partially effective.

Recent structural analysis of S1PRs has provided new insights into the plasticity of S1P lipid and its receptor binding mode (Chen et al., 2022; Yuan et al., 2021; Zhao et al., 2022). S1P binds to S1PRs with high affinity, but the actual binding between the ligand and individual S1PRs varies. For example, a large difference exists between the S1P‐bound S1PR₁ and S1P‐bound S1PR₃ in terms of ligand penetration and stabilisation and receptor recognition and activation, suggesting the dynamics of S1P‐bound receptors (Yu et al., 2022; Yuan et al., 2021). Both S1PR₁ and S1PR₅ possess a conserved ligand‐binding pocket consisting of a polar module and a hydrophobic cavity (Yuan et al., 2021). In contrast, S1PR₃ possesses multiple ligand‐binding pockets, indicating the plasticity of S1P‐S1PR interaction (Zeng et al., 2022). Differences also exist in the orthosteric sites between S1PR₁ and S1PR₃ (Zhao et al., 2022). In addition, S1P‐S1PR interaction can be influenced by chaperone proteins that bind and transport S1P in circulatory and interstitial fluid, facilitating S1PRs activation in a paracrine or endocrine manner (Cartier & Hla, 2019). However, different chaperone proteins such as apolipoprotein M and serum albumin bind to S1P at various affinities, thereby influencing the extent of S1PR activation on recipient cells. Collectively, these findings provide the structural basis demonstrating the selectivity and plasticity of ligand–receptor interaction between S1P and receptors, providing mechanistic insights into the biological roles of structural analogues of S1P that target individual S1PRs.

3.2. Crosstalk with other molecules

In addition to S1P, S1PRs are known to crosstalk with other structurally unrelated molecules such as activated protein C receptor, lysophosphatidic acid receptor‐1 (LPAR1), the hyaluronic acid receptor CD44, CD69, NogoA, and A2B Adenosine Receptors (A_2BR) (Coppi et al., 2020; Feistritzer & Riewald, 2005; Hisano et al., 2019; Kempf et al., 2014; Shiow et al., 2006; Studer et al., 2012); many of which have been implicated in myelination and remyelination (Kalafatakis et al., 2023; Miyamoto et al., 2016; Tuohy et al., 2004). A genome‐wide CRISPR/Cas9 screen identifies that LPAR1 negatively regulates S1PR₁ in the lymph node sinuses (Hisano et al., 2019). In the CNS, S1PR₂ acts as a receptor for NogoA (Kempf et al., 2014), a neuronal repulsive molecule that regulates neural plasticity. The extracellular domain of NogoA binds to S1PR₂, regulating myelin‐mediated inhibition of neurite outgrowth in the hippocampus and motor cortex via inducing the downstream G protein G₁₃‐RhoA pathway (Kempf et al., 2014). Whilst this primary effect of NogoA and S1PR₁ interaction lies in neurons, this finding could be therapeutically relevant as S1P modulators such as FTY720‐P can trigger G₁₃ activation via S1PR₂, although FTY720‐P is believed to activate S1PR_1,3,4,5. (Chen et al., 2022) and possibly binds to S1PR₂ with low affinity (Sobel et al., 2015). Directly relevant to oligodendrocyte development, crosstalk exists between the A_2BR and S1P signalling (Coppi et al., 2020). Reducing S1P production via blocking SphK1/2 prevents the increased expression of A_2BR during OPC differentiation (Coppi et al., 2020). Likewise, selectively silencing A_2BR in cultured OPCs prompts oligodendrocyte maturation, an effect associated with increased expression of SPL, which could conceivably result in reduced S1P levels (Coppi et al., 2020). Outside of the CNS, the crosstalk between S1P and other signalling pathways also exists in the Schwann cell myelination such as the Smad‐dependent tropism (Schira et al., 2019). Together, the crosstalk between S1PRs and other molecules could contribute to or interfere with S1P signalling in the process of oligodendroglial lineage progression and ultimately myelination, highlighting the complex interplay between S1P–S1PR interaction and the various cell types involved in the process of myelin formation.

3.3. Insights from current S1PR modulators to S1P signalling

Current S1P‐based therapies such as fingolimod and derivatives are functional analogues of S1P through agonising or antagonising S1PRs, functioning as S1PRs modulators (Chun et al., 2021). However, not all biological effects between these S1PR modulators and endogenous S1P reconcile. S1P protects against neurotoxicity, however, this protective effect is not observed with fingolimod (Tran et al., 2020). Fingolimod resembles the chemical structure of sphingosine. It is phosphorylated by the sphingosine kinase 2 to become the active form, fingolimod‐phosphate (fingolimod‐P) (Brinkmann et al., 2002). In a murine model of neuron–glia cocultures, S1P but not fingolimod‐P protects against the death of hippocampal neurons, an effect driven by S1PR₂‐induced transcription in astrocytes (Tran et al., 2020). In primary astrocyte cultures, the addition of S1P drastically up‐regulates an array of neurotrophic factors including the brain‐derived neurotrophic factor, leukaemia inhibitory factor, platelet‐derived growth factor (PDGF), and heparin‐binding EGF‐like growth factor, whereas fingolimod‐P exerts a marginal effect (Tran et al., 2020). Similarly, S1P increases extracellular glutamate in cultured astrocytes via signalling to S1PR₂, an effect not observed with siponimod or fingolimod (Jonnalagadda et al., 2021). It is well established that S1P binds to five S1PRs (S1PR_1–5) with high affinity, whilst fingolimod‐P binds to four S1PRs (S1PR_1,3,4,5) (Kihara et al., 2014; Mandala et al., 2002), explaining the relative inefficacy of fingolimod‐P compared with S1P. Moreover, elevating S1P levels through genetically or pharmacologically blocking its degradation demonstrates a protective effect against myocardial infarction (Bandhuvula et al., 2011), which also cannot be achieved through modulating S1PRs (Knapp, 2011). Mechanically, the differential effects observed between endogenous S1P and S1PR modulators are almost attributable to the complexity of S1P signalling mediated through coupling proteins and multiple receptors, potentially leading to differential downstream signals.

Current S1PR modulators such as fingolimod and its derivatives interact with G proteins and S1PRs, in a manner different from endogenous S1P (Yuan et al., 2021; Zeng et al., 2022). S1P induces receptor internalisation and subsequent endosomal recycling, whereas fingolimod and fingolimod‐P induce irreversible receptor internalisation and degradation (Cyster & Schwab, 2012; Rosen et al., 2013). Fingolimod and its derivatives act as non‐selective modulators of S1PR that possess dual agonism and antagonism properties (Huwiler & Zangemeister‐Wittke, 2018; Imeri et al., 2021; Stepanovska Tanturovska et al., 2021). Fingolimod not only binds and activates four S1PRs (S1PR_1,3,4,5) but also functionally antagonises S1PR₁ (Zeng et al., 2022). The specificity of fingolimod derivatives also varies. Like the parent compound, an oxazolo‐oxazolo derivative (ST‐1071) of fingolimod internalises S1PR₁, induces lymphopenia and reduces the clinical severity of EAE in mice (Imeri et al., 2014). Modification of ST‐1071 leads to the identification of new derivatives ST‐1505 and ST‐1478 which exert similar effects in EAE mice (Imeri et al., 2021). However, mechanistically, ST‐1505 is characterised as a dual S1PR_1,3 agonist, whereas ST‐1478 is a dual S1PR_1,5 agonist (Imeri et al., 2021). A bismorpholino derivative of oxy‐fingolimod (ST‐2191) processes dual potency on S1PR₁ agonism and functional antagonism, although it is proven effective in reducing lymphocytes in mice (Stepanovska Tanturovska et al., 2021). Therefore, whilst these fingolimod derivatives are considered potent S1PR₁ modulators that ameliorate the EAE model in mice via inducing lymphopenia, they display a diverse profile of specificity. Moreover, other S1PR modulators such as ozanimod and etrasimod are small molecules that agonise S1PR_1,5 and S1PR_1,4,5, respectively (Sandborn et al., 2016), in a manner similar to fingolimod (Cahalan et al., 2011; Marchese & Benovic, 2004). Most cells express two or more subtypes of S1P receptors (Chun et al., 2019) and each S1P receptor can form a receptor complex that couples with distinct G proteins, adding another layer of complexity to S1P signalling. Whilst S1PR₁ is the most widely expressed receptor and is found ubiquitously throughout various tissues, S1PR₅ is prominently found in the CNS (Fang et al., 2021; Green et al., 2021). Different CNS cells may possess different S1PR subtypes in varying densities and thus help to modulate the S1P‐S1PR axis through differential pathways, controlling various activities within the CNS. Therefore, the development of therapeutics that target S1PRs could be hindered by the high sequence homology among these receptors and the intricate nature of S1P–S1PR interactions. The effects of current S1PR modulators on S1P signalling remain controversial and do not fully substitute endogenous S1P and its function.

4. S1P REGULATES CNS MYELIN FORMATION IN THE CNS

4.1. Biosynthesis of S1P: Pro‐survival of OLG?

The role of S1P in OLG and CNS myelination has been investigated using in vitro and in vivo models, with supporting evidence shedding light on a promyelinating effect. In vitro, exogenous S1P appears to exert stage‐specific effects on OLG via signalling to distinct intracellular pathways, inducing process retraction of immature OLG via the Rho‐Rho‐associated protein kinase (Rock) pathway whilst promoting the survival of mature OLG via an Akt‐dependent pathway (Jaillard et al., 2005). S1P also influences other intracellular signalling cascades involved in oligodendrocyte development such as the Ras/Raf/mitogen‐activated protein kinases (MAPK)/ERK pathway. Specifically, S1P transiently increases the phosphorylation level of ERK in differentiated rat cortical OLG in vitro (Yu et al., 2004), an effect completely abolished by MAPK Inhibitors. MAPK/ERK pathway is an essential intracellular pathway regulating myelin formation in vitro (Gonsalvez et al., 2016; Xiao et al., 2012) and in vivo (Ishii et al., 2012, 2014, 2019). Interestingly, S1P exerts no effect on intracellular calcium levels in the cultured cortical OLG (Yu et al., 2004). Genetic analysis shows that calcium attenuation in OLG does not affect the overall number of myelinated axons or myelin thickness, but instead, impedes the longitudinal extension of myelin sheaths (Iyer et al., 2024). Therefore, these findings collectively suggest that S1P could promote oligodendrocyte survival to potentiate myelin generation, presumably by regulating the canonical intracellular signalling cascade such as ERK/MAPK.

Whilst the exact mechanism remains unclear, the role of S1P in promoting oligodendrocyte survival indicates a beneficial effect in MS by protecting OLG from cell death in myelin lesions in vivo. Delivery of exogenous S1P in vivo is challenging due to binding by carrier proteins and rapid metabolism. One strategy to alter the level of endogenous S1P is to regulate SphK enzymes (SphK1 and SphK2), which synthesise S1P by phosphorylating sphingosine, thereby influencing its biological functions (Diaz Escarcega et al., 2021). Blondeau et al. (2007) demonstrate that SphK2 is the predominant isoform found in neural parenchymal tissues. Analysis of the SphK2 KO mouse indicates that endogenous S1P potentiates oligodendrocyte survival (Lei et al., 2019) and myelin formation in vivo (Song et al., 2021), although an effect during development is not reported. It is unclear if SphK1 plays a role in regulating myelin formation or compensates for SphK2's effect as the varying presence of the SphK1/2 could compete with S1P production in response to the ablation of one SphK isoform. The SphK1 enzyme subtype differs from the SphK2 since it is predominantly found in the cytoplasm, whilst SphK2 resides within cell organelles including the nucleus, endoplasmic reticulum, and mitochondria (Diaz Escarcega et al., 2021). Although both enzyme isoforms catalyse sphingosine phosphorylation with the same mechanism, the varying presence of the two subtypes across the cellular and tissue levels across different phases of development is indicative of the differential physiological implications (Maceyka et al., 2005). Findings by Maceyka et al. (2005) demonstrate that the involvement of either subtype may be able to alter the effects of S1P produced by their actions on the receptor signalling pathway, with opposing effects. Although the level of SphK1 is not assessed in the SphK2 KO mouse brain as a result of technical issues (Song et al., 2021), it is possible that increased S1P in SphK2 KO mice could lead to high SphK1 activity in CNS cells (Fischer et al., 2011; Nayak et al., 2010). Collectively, these are important findings and indicate that endogenous S1P potentiates oligodendrocyte survival and myelination. Adopting a SphK1/2 double KO strategy would help dissect the role of endogenous S1P in CNS myelination in the future.

4.2. S1P‐S1PR signalling in oligodendroglia: A stage‐specific effect?

OLG express S1PR₅ as well as S1PR₁ and S1PR₂ (Jaillard et al., 2005; Novgorodov et al., 2007; Yu et al., 2004), implying a role of S1P signalling in myelination (Figure 1b), although limited evidence is available to reveal a myelin phenotype in vivo. OLG express S1PR₁ and S1PR₅, the latter of which is expressed throughout the lineage development from OPCs to mature myelin‐forming cells (Jaillard et al., 2005). Interestingly, S1PRs display distinct responses to mitogens involved in oligodendrocyte function. For example, PDGF, a pro‐survival/proliferative factor of OLG, differentially regulates the expression of S1PRs, down‐regulating S1PR₅ in OPCs whilst up‐regulating S1PR₁ (Jung et al., 2007). This finding indicates that S1PR₅ could play a prominent role in postmitotic OLG whilst the function of S1PR₁ residuals within proliferating OPCs. Indeed, in rat cortical OPC cultures, the application of FTY720P, an S1PR₁ modulator, improves OPC survival after serum withdrawal (Jung et al., 2007), suggesting a role of S1PR₁ in potentiating OPC survival or proliferation. In vitro, S1P decreases OPC migration, an effect attenuated by the siRNA‐based knocking down of S1PR₅, but not S1PR₂, implying that S1PR₂ does not play a major role in regulating S1P's effect on OLG (Novgorodov et al., 2007). S1PR₅ is considered the primary target of S1P signalling in OLG as a result of its abundant and almost exclusive expression profile primarily restricted to differentiated OLG (Jaillard et al., 2005; Yu et al., 2004). Within the oligodendroglial lineage, S1PR₅ exhibits a dual‐signalling role: inducing process retraction in pre‐myelinating OLG via the Rho kinase/CRMP2 pathway, whilst promoting the survival of mature OLG through the Akt signalling pathway (Jaillard et al., 2005). Therefore, the above findings suggest that S1PRs serve different functions during the development of oligodendroglial lineage cells and that S1P could exert stage‐specific effects on OLG depending on the receptor being activated: promoting the survival/proliferation of OPCs via activating S1PR₁ and potentiating the survival/differentiation of differentiated OLG via S1PR₅. This multifaceted function underscores the significance of S1P signalling in oligodendrocyte function, particularly in demyelinating diseases such as MS.

The broad spectrum of S1PR expression in CNS cells implies that S1P could influence myelin formation via cell–cell interaction during neural development. S1PR₁ is known to regulate neurogenesis and angiogenesis within the CNS (Blaho & Hla, 2014; Jiang et al., 2023). However, there is a lack of literature demonstrating if the loss of S1PR₁ impedes the development of OLG or myelin formation in vivo. Analysis of the S1PR₂ mouse mutant together with a pharmacological approach demonstrates that S1PR₂ regulates the BBB integrity particularly permeability, ultimately contributing to CNS demyelination (Cruz‐Orengo et al., 2014). S1PR₃ is implicated in astrogliosis (Dusaban et al., 2017) and microglia‐mediated inflammation (Gaire et al., 2018). S1PR₄ inhibits cell proliferation in vitro (Graler et al., 2003). Upon S1P stimulation, S1PR₄ activates the small GTPase Rho and undergoes cytoskeletal rearrangements, inducing peripheral stress fibre formation and cell rounding (Graler et al., 2003). In CNS, the function of S1PR₄ is most active in dendritic cells instead of a direct influence upon OLG, although it could alter extracellular cues via regulating the secretion of effector cytokines (Graler et al., 2003).

Taken together, the precise role of S1P in regulating oligodendrocyte function and myelin formation is largely unclear, primarily because of the limited evaluation of S1P in myelination via modulating its biosynthesis or downstream signalling using genetic or pharmacological approaches. Whether S1P signalling plays a role in regulating OLG and maintaining myelin integrity in the adult CNS remains another unexplored topic. The unclear function of individual S1PRs in OLG complicates the interpretation of current S1P modulators that target these receptors for treating demyelinating diseases. Despite these, recent studies provide valuable insights into S1P's promyelinating effects in the mammalian CNS.

5. THE ROLE OF S1P IN MODULATING MYELIN REPAIR AFTER CNS INJURY

5.1. S1P biosynthesis regulates myelin damage and repair

The sphingolipid family upstream of S1P production is known to regulate myelinogenesis and is involved in demyelinating disorders in developmental and acquired conditions (Giussani et al., 2021; Jana & Pahan, 2010). Sphingomyelin is essential to oligodendrocyte survival and myelin sheath formation (Yoo et al., 2020). Moreover, ceramide levels correlate with the extent of myelinogenesis (Majumder et al., 2020). Interrupting the synthesis of very long‐chain ceramide through ceramide synthase 2 (CerS2) ablation in OLG triggers microglial activation and myelin degeneration in multiple CNS regions including the corpus callosum, cerebellum, and striatum, as evidenced by a drastic reduction of myelin protein expression and myelin thickness independent of axonal diameter (Teo et al., 2023). Importantly, this myelin structural change, as a result of very long‐chain ceramide deletion, is accompanied by motor deficits. Collectively, these findings highlight the importance of the sphingolipid family in regulating myelin integrity after CNS injury.

The level of S1P is reduced during the peak of demyelination in vivo (Kim et al., 2018). Studies targeting the biosynthesis of S1P or downstream signalling imply that S1P regulates myelin damage and repair after CNS injury. Analysis of the SphK2 KO mouse shows that reducing S1P tissue levels cooperates with an amyloidogenic transgene and induces oligodendrocyte loss and demyelination in a mouse model of Alzheimer's disease (Lei et al., 2019). Interestingly, reducing S1P level does not alter OPC proliferation and differentiation as assessed by the number of OPCs and mature OLG in the cuprizone model of central demyelination, but significantly impedes the extent of remyelination for at least 4 weeks after cuprizone withdrawal as assessed by myelin marker protein expression (Song et al., 2021). Whilst studies that investigate S1P biosynthesis in the context of myelination are limited, the above findings imply that endogenous S1P is required for the remyelinating process via potentiating the survival of OLG in myelin lesions. Future research is required to investigate whether elevating S1P tissue levels or stimulating S1P signalling can rescue oligodendrocyte loss in SphK2 KO mice or restore remyelination following cuprizone withdrawal. Several inhibitors of SPL have been proven effective in elevating S1P levels in vivo (Bagdanoff et al., 2009; George & Xiao, 2024; Kim et al., 2013; Ohtoyo et al., 2016; Schwab et al., 2005; Weiler et al., 2014). Evaluating their role in oligodendroglial cells will provide new insights into the mechanism by which endogenous S1P regulates myelin formation and repair.

5.2. S1P–S1PR signalling regulates myelin damage

S1PRs display different responses to a demyelinating insult in vivo. Analysis of a murine demyelinating model shows that the expression of S1PR₁, S1PR₃, and S1PR₅ increases during the progression of demyelination, whilst S1PR₂ expression remains unchanged (Kim et al., 2018), suggesting distinct roles of S1PRs in myelination following CNS injury.

5.2.1. S1PR₁

Several studies investigating S1PR mutant mice or partial S1P analogues demonstrate that S1P–S1PR₁ signalling regulates myelination after injury in vivo. In the cuprizone model of central demyelination, conditional deletion of S1PR₁ specifically in oligodendroglial lineage cells (S1PR₁ ^fl/fl CNP^Cre) in mice leads to significantly fewer myelinated axons and CC1+ mature OLG compared with wild‐type control mice (Kim et al., 2011), suggesting that endogenous S1P is required to protect oligodendrocyte and myelin loss at the basal level. Consistent with this finding, FTY720 (fingolimod, a functional analogue of S1P) and CYM5442 (a short‐lived S1PR₁ modulator) exert a similar effect on ameliorating demyelination in the cuprizone‐induced demyelination model, assessed by myelin intensity and the number of myelinated axons (Kim et al., 2011, 2018). FTY720's protective effect against cuprizone‐induced demyelination is accompanied by an increase in the number of proliferating OPCs (NG2 + PCNA+) and late progenitor cells (Nkx2.2+), suggesting that S1P signalling protects oligodendroglia lineage cells in myelin lesions (Kim et al., 2011). Interestingly, the administration of FTY720 or CYM5442 drastically reduces S1PR₁ expression, but not the rest of S1PRs, and prevents oligodendrocyte death accompanied by an overall reduction of astrogliosis, microglia accumulation and proinflammatory mediators (Kim et al., 2011, 2018). These findings imply that S1P–S1PR₁ signalling could protect OLG in myelin lesions by suppressing neuroinflammation (Figure 1b). Similar to the cuprizone model findings (Kim et al., 2011, 2018), FTY720 and SEW2871 (an S1PR₁ agonist) inhibit lysolecithin‐induced demyelination as assessed by myelin marker protein expression (Sheridan & Dev, 2012). The S1PR₁‐regulated myelin protection likely results from receptor agonism rather than functional antagonism, since mice lacking S1PR₁ in OLG exhibit significantly worse demyelination in the cuprizone model (Kim et al., 2011). Taken together, these studies provide convincing evidence that S1P–S1PR₁ could protect demyelination presumably by promoting the survival of OPCs.

However, the role of S1PR₁ in remyelination remains to be determined. Whilst in vitro or ex vivo studies identify that FTY720 stimulates remyelination in cerebellar slices (Miron et al., 2010), it fails to enhance spontaneous remyelination in in vivo animal studies (Alme et al., 2015; Hu et al., 2011; Nystad et al., 2020). Adopting a remyelinating paradigm (after cuprizone withdrawal), FTY720 and CYM5442 exert no effect on remyelination in the cuprizone model of central demyelination (Kim et al., 2011, 2018). In a separate study adopting a similar therapeutic paradigm, FTY720 exhibits no effect on the number of mature OLG and microglia as well as astrocytes in the cerebellum (Alme et al., 2015), although FTY720 administration leads to more OLG in the secondary motor cortex at 3 weeks of remyelination (Nystad et al., 2020). In a demyelination model induced by the mouse hepatitis virus, administration of FTY720 enhances the migration and proliferation of transplanted neural progenitor cells but again exerts no effect on the extent of myelin repair (Blanc et al., 2015). Collectively, current evidence supports the role of S1PR₁ activation in protecting OPCs and demyelination following injury in the mammalian CNS; however, its role in remyelination is yet to be determined.

5.2.2. S1PR₂

In contrast to the protective effect of S1PR₁ on myelin, S1P signalling via S1PR₂ could be detrimental to myelin integrity after injury (Figure 1b). Mice with S1PR₂ deficiency exhibit less demyelination and fewer inflammatory foci during the peak of the EAE model of inflammatory central demyelination compared with control mice (Cruz‐Orengo et al., 2014), suggesting that activating S1P signalling via S1PR₂ accelerates demyelination. S1PR₂ is known to regulate vascular biology. In the CNS, S1PR₂ is primarily expressed by endothelial cells and vessel‐associated astrocytes as well as occasional pericytes (Cruz‐Orengo et al., 2014), all of which are associated with BBB integrity. Analysis of the S1PR₂ mouse mutant demonstrates that S1PR₂ destabilises adherence junctions, thereby altering BBB permeability (Cruz‐Orengo et al., 2014). Inhibiting S1PR₂ via a specific inhibitor JTE‐013 or S1PR₂ ablation preserves BBB polarity and ameliorates EAE disease severity (Cruz‐Orengo et al., 2014). In a model of retinal ischemia, inhibiting S1PR₂ via JTE‐013 reverses the extent of optic nerve demyelination, as assessed by myelin intensity and the thickness of myelin membranes (via G ratio) (Xue et al., 2023). This myelin phenotype is accompanied by preserved visual function and more mature OLG (CC1+) and remyelinated axons in the JTE‐013‐treated ischemia model compared with control mice (Xue et al., 2023). Taken together, these findings suggest that S1P–S1PR₂ signalling induces demyelination in vivo via regulating glial cell function involved in BBB permeability.

5.2.3. S1PR₅

S1PR₅ has been reported to regulate the pro‐survival effect of S1P in mature OLG in vitro (Jaillard et al., 2005) (Figure 1b). Like S1PR₁, analysis of S1PR₅ KO suggests its expression is not required for oligodendrocyte survival and the extent of demyelination at the basal level in vivo (Behrangi et al., 2022). Siponimod is an S1PR_1/5 dual agonist initially identified via a Xenopus‐based drug screen (Mannioui et al., 2018). A subsequent Crispr/cas9 gene editing approach confirms that siponimod's promyelinating effect requires the expression of S1PR₅ (Mannioui et al., 2018). In the cuprizone model of central demyelination, siponimod ameliorates the loss of OLG and myelin, an effect abolished in S1PR₅ KO mice (Behrangi et al., 2022). These results not only indicate that siponimod exerts a direct effect on OLG but also identify S1PR₅ is a key player that regulates S1P‐induced effect on myelin protection in vivo. Whilst no reports suggest S1PR₅ regulates oligodendrocyte remyelination in the mammalian CNS, a Xenopus oocytes demyelination model identifies that S1PR₅ deletion impedes remyelination mediated by siponimod (Mannioui et al., 2018). Data from an in vitro spheroid cell myelinating model shows that FTY720 augments myelin markers during a remyelination phase primarily by S1PR₅ agonism in addition to suppressing microglial activity and protecting oligodendrocyte apoptosis (Jackson et al., 2011). Taken together, current findings suggest that S1P signalling plays a promising role in protecting OLG and myelin after a demyelinating insult, protecting OPCs and suppressing neural inflammation via S1PR₁ whilst promoting oligodendrocyte survival and protecting myelin loss via S1PR₅ (Figure 1b).

6. CONCLUSION AND PERSPECTIVES

Pharmacological and genetic evidence identify S1P as a pivotal regulator of OLG and myelination. The amount of S1P in the CNS exceeds that found in other peripheral organs, such as the liver or spleen (Edsall & Spiegel, 1999), indicating its important implication in regulating CNS function. In the mammalian CNS, stimulating S1P biosynthesis demonstrates that endogenous S1P prevents demyelination and promotes remyelination, however, agonism of S1P receptors using partial S1P ligand or S1PR modulators protects myelin loss but does not promote myelin repair. Albeit, to what extent endogenous S1P regulates remyelination in vivo remains opaque in the mammalian CNS. That said, S1P signalling potentiates peripheral nerve remyelination as its expression is essential to myelinogenesis after peripheral nerve injury (Meyer Zu Reckendorf et al., 2020). Hence, it remains interesting to investigate if a similar pathway downstream of S1P is involved in CNS remyelination as in the PNS. Remyelination failure in central demyelinating diseases such as MS is complex. It is thought to be because of the limited availability of OLG, stalled survival and differentiation of OLG, and the presence of inhibitors in inflamed myelin lesions (Mitew et al., 2014). Although the precise role that S1P plays in CNS myelination is not fully understood, the trophic role that S1P signalling plays in OPCs and mature OLG shows a significant potential of S1P in promoting remyelination via enabling the supply of residual OLG in myelin lesions.

It is evident that targeting individual S1PRs cannot fully recapitulate the biological role of endogenous S1P. These differential effects are almost attributable to the complex and dynamic interplay between S1P and five S1P receptors as well as the associated G proteins, which together contribute to the execution of S1P signalling in various cell types involved in the myelinating process. The lack of sophisticated genetic tools that modulate S1P biosynthesis or S1PR activity poses additional challenges. The conflicting evidence between various model systems indicates that S1P–S1PR signalling regulates de/remyelination via cell–cell interaction rather than a sole influence upon OLG since demyelination in in vitro or ex vivo models is induced chemically and excludes possible effects of immune or vascular cells present in in vivo settings (Alme et al., 2015; Hu et al., 2011; Nystad et al., 2020).

In summary, S1P represents a promising target for promoting CNS myelination and holds significant therapeutic potential for addressing demyelinating diseases and neurological disorders. Several key questions remain to be addressed. The precise mechanism by which S1P–S1PR signalling regulates oligodendrocyte development and myelin formation is unclear. S1P is known to regulate CNS homeostasis; however, it remains unknown whether S1P is involved in maintaining the homeostasis of OLG and myelin in the mature and ageing CNS. Continued efforts aimed at unravelling the complexity of S1P signalling pathways and translating these findings into clinical applications will pave the way for improving therapeutic outcomes for patients living with demyelinating diseases such as MS. In human MS brains, the level of S1P is lower in MS lesions compared with the normal‐appearing white matter (Qin et al., 2010), implying a role of S1P in MS lesions. However, the delivery of exogenous S1P lipids is challenging as a result of binding by carrier proteins in recipient cells and rapid metabolism, arguing the need to explore new strategies that selectively control S1P biosynthesis or degradation in the sphingolipid metabolic network. Such strategies will be critical to decipher the role of S1P in remyelinating and to aid the development of future new S1P‐based therapeutics.

AUTHOR CONTRIBUTIONS

Fatima Binish: Data curation; methodology; writing – original draft; writing – review and editing. Junhua Xiao: Conceptualization; data curation; funding acquisition; investigation; methodology; project administration; writing – original draft; writing – review and editing.

FUNDING INFORMATION

The work presented in the manuscript is supported by the Judith Jane Mason and Harold Stannett Williams Memorial Foundation National Medical Program (#Mason2210).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of financial or non‐financial interests that are directly or indirectly related to the work submitted for publication.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/jnc.16228.

Binish, F. , & Xiao, J. (2025). Deciphering the role of sphingosine 1‐phosphate in central nervous system myelination and repair. Journal of Neurochemistry, 169, e16228. 10.1111/jnc.16228

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.