The Contributions of the mTOR Complexes: How Does Regional and Temporal Heterogeneity Affect Myelination and Remyelination?

Myelin, generated by oligodendrocytes (OLs), is essential to the function of the CNS as it insulates axons and provides metabolic support (Stadelmann et al., 2019). Myelination, the process by which axons are ensheathed by OLs, begins during development and persists throughout adulthood, as oligodendrocyte precursor cells (OPCs) differentiate into mature myelinating OLs (Bergles and Richardson, 2015). Demyelination contributes to dysfunction in pathologies, such as multiple sclerosis (MS) (Höftberger and Lassmann, 2018) and spinal cord injury (SCI) (Waxman, 1989). Hence, OPC proliferation and differentiation into new OLs and remyelination are crucial processes for recovery from MS and SCI (Huang and Franklin, 2011; Gaesser and Fyffe-Maricich, 2016; Duncan et al., 2020), helping to restore electrophysiological function and mitigate secondary degeneration (Pang et al., 2014). Remyelination and developmental myelination share some common mechanisms (Pang et al., 2014); for example, extracellular signal-regulated kinase 1/2 is implicated in both development (Ishii et al., 2013) and remyelination after SCI (Fyffe-Maricich et al., 2013). Acquiring a fuller mechanistic understanding of developmental myelination may therefore guide the development of remyelination treatments (Pang et al., 2014).

With injury and disease, OPC physiology and responsiveness to differentiation cues change (Bergles and Richardson, 2015), and the cells can acquire detrimental as well as beneficial effects. For example, OPCs' capacity as multipotent progenitors increases after SCI, so that OPCs differentiate not only into myelinating OLs but also into astrocytes that contribute to formation of the glial scar that restricts axon regeneration (Duncan et al., 2020). In MS, a reduced differentiation ability of OPCs contributes to impaired myelin repair (Huang and Franklin, 2011), and proliferating OPCs contribute to inflammation and subsequent neurodegeneration in the demyelinated lesion (Psenicka et al., 2021). Understanding and distinguishing the mechanisms that promote OPCs' differentiation to remyelinating OLs versus differentiation to astrocytes or promotion of inflammation are important, as they elucidate possible targets for remyelination efforts.

In the intact, injured, and diseased CNS, OPCs pass through multiple transitionary stages to differentiate into myelinating OLs: proliferative OPCs, pre-OLs, immature differentiated OLs, and mature myelinating OLs (Tiane et al., 2019). Progression of OL-lineage cells through these stages is driven by stage-specific alterations of signaling pathways (Adams et al., 2021) that coordinate protein and lipid synthesis and promote cytoskeleton reorganization (Figlia et al., 2018; Brown and Macklin, 2019). Cytoskeletal dynamics are important at multiple stages. For example, actin polymerization drives early OPC differentiation and the extension of OPCs processes toward axons for ensheathment, while actin depolymerization drives myelin compaction (Brown and Macklin, 2019). Thus, actin polymerizing regulators are enriched in differentiating and premyelinating OLs, whereas actin depolymerizing regulators (e.g. gelsolin) are enriched in myelinating OLs (Thomason et al., 2020).

One important regulator of OPC differentiation and myelination during development is the mechanistic target of rapamycin (mTOR) (Figlia et al., 2018). mTOR is a serine/threonine kinase that makes up two distinct complexes, mTOR complex 1 (mTORC1) and mTORC2, defined by the presence of Raptor and Rictor, respectively (Figlia et al., 2018). Across cell types, mTORC1 signaling primarily regulates cell growth and metabolism, while mTORC2 regulates cell proliferation, survival, and cytoskeletal organization (Saxton and Sabatini, 2017). More specifically, mTORC1 signaling is required for late-stage (OL maturation to myelinating OL) development and myelination in the spinal cord, but not in the corpus callosum (Bercury et al., 2014; Lebrun-Julien et al., 2014). In contrast, mTORC2 seems necessary for early-stage (OPC to immature OL) differentiation in the corpus callosum. Moreover, its ablation in brain OPCs results in a recoverable delay in myelination, suggesting that mTORC2 has a transient role in myelination (Grier et al., 2017). Whether mTORC2 regulates early-stage OPC differentiation and myelination in the spinal cord is unclear, however.

To clarify the roles of mTORC2 in OL lineage cells, Dahl et al. (2023) conditionally deleted Rictor, a necessary component of the mTORC2 complex (Jebali and Dumaz, 2018), in OPCs in mice, using PDGFRα-Cre X Rictor^fl/fl mice (Rictor cKO). They investigated the effects of Rictor deletion on early-stage OPC differentiation in both the corpus callosum and the spinal cord. Specifically, they examined the impact of Rictor loss on the numbers of oligo-lineage cells, levels of myelin protein and RNA, myelin morphology, signaling by mTORC2 and mTORC1, and actin regulators. Their data suggest that mTORC2 signaling loss results in significant reduction in OPC differentiation and in long-term hypomyelination in the corpus callosum, with minimal effects in the spinal cord.

Compared with controls, the Rictor cKO corpus callosum had significantly fewer oligo-lineage cells (Olig²⁺ cells/mm³) and differentiated OLs (CC1⁺ cells/mm³), and a lower ratio of differentiated OLs to total OL-lineage cells at postnatal day 14 (P14) and P30, suggesting reduced OPC differentiation (Dahl et al., 2023, their Fig. 2C-E). Myelin-associated glycoprotein, 2′,3′-cyclic nucleotide 3′-phosphodiesterase, myelin basic protein, and myelin oligodendrocyte glycoprotein levels were significantly reduced in Rictor cKO corpus callosum at P14 and P30, and myelin proteolipid protein levels were reduced at P14 (Dahl et al., 2023, their Fig. 3A′–E′). Reduction in proteins persisted to P350 (Dahl et al., 2023, their Fig. 8G), suggesting that mTORC2 has a long-term role in myelination. In addition, Rictor cKO mice had significantly reduced myelinated axon density than control mice at P45 (Dahl et al., 2023, their Fig. 7B). Moreover, the proportion of myelinated small-diameter axons was larger in Rictor cKO mice than in controls (Dahl et al., 2023, their Fig. 7C), suggesting that mTORC2 disruption preferentially affects myelination of larger axons in early-stage OL differentiation. However, myelin morphology, including relative myelin thickness measured by g-ratio (the ratio of axon to total fiber diameter), was similar to that of control mice at P21 and P45 (Dahl et al., 2023, their Fig. 7C).

Because cytoskeletal reorganization is influenced by the mTOR pathways (Oh and Jacinto, 2011) and is important in OPC differentiation and myelination (Brown and Macklin, 2019), Dahl et al. (2023) investigated actin polymerization and depolymerization regulators in the Rictor cKO corpus callosum. Specifically, profilin (involved in polymerization of G-actin into F-actin) and gelsolin (involved in Ca²⁺-dependent actin filament severing and depolymerization) levels were examined (Thomason et al., 2020). In Rictor cKO mice, profilin1 levels were reduced only at P14 (Dahl et al., 2023, their Fig. 9H), but gelsolin levels were reduced at all stages from P14 persisting to P350 (Dahl et al., 2023, their Fig. 9F). Because gelsolin is most highly expressed in actively myelinating OLs (Thomason et al., 2020), this result suggests a role for mTORC2 in corpus callosum myelination.

Given that mTORC2 and mTORC1 signaling is interconnected (mTORC2 activation leads to mTORC1 activation) (Polchi et al., 2018), Dahl et al. (2023) investigated whether Rictor deletion alters levels of the key targets of mTORC2 and mTORC1 in the corpus callosum. Rictor cKO significantly reduced phosphorylation of its effectors Akt and PKCα/β at P7, P14, and P30 (Dahl et al., 2023, their Fig. 4B,C), indicating reduced mTORC2 signaling. In contrast, Rictor cKO did not alter phosphorylation of mTORC1 effectors. However, at P30, phosphorylation of the mTORC1 effector S6RP was reduced in the corpus callosum of Rictor cKO mice (Dahl et al., 2023, their Fig. 4E), likely because of reduced numbers of differentiating OLs. Together, these data suggest that, although Rictor loss does not initially affect mTORC1 signaling, it does have a small impact on mTORC1 signaling by P30.

In contrast to effects in the corpus callosum, Dahl et al. (2023) found that Rictor loss had minimal effects on oligo-lineage cells in the spinal cord. Specifically, Rictor cKO did not significantly change OPC differentiation at P14 and P30 (Dahl et al., 2023, their Fig. 2F-H), myelination at P30 (Dahl et al., 2023, their Fig. 1D), or density or morphology of individual myelinated axons at P21 and P45 (Dahl et al., 2023, their Fig. 7D-F) in the spinal cord. Nevertheless, Rictor loss had limited effects on myelin-related protein and RNA expression in the spinal cord at P14 and P30 (Dahl et al., 2023, their Fig. 5). Moreover, Akt phosphorylation by mTORC2 was reduced in the spinal cord of Rictor cKO mice at P7, P14, and P30 in Dahl et al. (2023, their Fig. 6A), and profilin1 and profilin2 levels were reduced at P14 and P30, respectively (Dahl et al., 2023, their Fig. 9I,K), suggesting that mTORC2 makes some contributions to cytoskeleton regulation in the spinal cord. These findings suggest that OL differentiation and myelination in the spinal cord do not depend heavily on mTORC2.

Notably, although PKCα was present in the spinal cord, there was no evidence of PKCα/β phosphorylation by mTORC2 in the spinal cord of either Rictor cKO or control mice during development (Dahl et al., 2023, their Fig. 6B). This suggests that pPKCα/β does not play a role in developmental myelination in the spinal cord. The absence of pPKCα/β in the control spinal cord, presence of pPKCα/β in control corpus callosum, and reduced pPKCα/β in the Rictor cKO corpus callosum suggest that pPKCα/β may account for the regional difference in mTORC2-dependent regulation of OPC differentiation (Dahl et al., 2023, their Figs. 4C, 6B).

Overall, Dahl et al.'s (2023) paper contributes to the understanding of the temporal and spatial heterogeneity of OPCs (for review, see Sherafat et al., 2021). They identify the key roles of mTORC2 in early-stage OL development in the corpus callosum, contrasting mTORC1's roles at late-stage OL differentiation and myelination in the spinal cord, but not the corpus callosum (Bercury et al., 2014; Lebrun-Julien et al., 2014).

As our understanding of mTORC2 and mTORC1 in the developing CNS grows, we should investigate the role these pathways play in reestablishing lost myelin after disease or injury. Some experiments have broadly targeted mTOR signaling to study its effects on remyelination. For example, mTOR loss from OPCs in rodent cuprizone-induced demyelination models compromised metabolic function in differentiating OLs and delayed myelin production in the brain, highlighting the contributions of mTOR pathway to remyelination (Jeffries et al., 2021). To our knowledge, no studies have specifically manipulated mTORC1 or mTORC2 activity in the oligo-lineage to investigate their role on OPC differentiation in MS or SCI. Given the contributions of mTORC2 to developmental myelination in the brain (Dahl et al., 2023), and mTORC1 in the spinal cord (Bercury et al., 2014), these pathways may have potential as targeted remyelination therapies, specific to the pathologic region. MS afflicts the entire CNS; thus, oligo-lineage-specific targeting of mTORC1 and mTORC2 signaling could represent a promising area of future research. Likewise, targeting mTORC1 signaling may have therapeutic potential after SCI. Finally, considering the temporally specific roles of these pathways, future research should investigate their therapeutic potential to target specific stages of OL differentiation. Overall, the mTORC1 and mTORC2 pathways may prove beneficial to developing targeted remyelination therapy.