The role of TSC1 and TSC2 proteins in neuronal axons

3.1. TSC/mTOR axis in translation

TSC1/2 has been canonically considered to control protein synthesis primarily by modulating the function of mTORC1⁷⁵. mTORC1 was identified as a positive translational regulator in neurons in the early 2000s when it was first shown to be required for the de novo protein synthesis during the late phase of hippocampal long-term potentiation (LTP)⁷⁶. The best characterized mTORC1 downstream targets through which the complex controls translation are the ribosomal protein S6 kinase beta-1 (S6K1/p70S6K1) and 4E binding proteins (4E-BPs)⁷⁷ (Fig. 1). While it has been widely accepted that TSC-mediated protein synthesis is mTOR-dependent, some studies suggest that TSC1/2 or Rheb can potentially regulate translation in mTOR-independent ways as well. For example, in response to cell stress, Rheb can bind directly to protein kinase-like endoplasmic reticulum kinase (PERK), induce PERK-mediated-phosphorylation of the eukaryotic translation initiation factor a2 (eIF2a), and arrest protein synthesis in HEK293 cells⁷⁸. Notably, TSC1/2-dependent effects on the translatome exhibit variable sensitivity to rapamycin treatment⁷⁹. While this observation may be a result of either differential turnover timelines of these mRNAs and proteins or different treatment paradigms, it further supports the notion that TSC1/2 may regulate translation independently of mTORC1.

Several models have been utilized to elucidate the contribution of TSC1/2-dependent translational changes to TSC phenotypes, including mouse models and human induced pluripotent stem cells (iPSCs). Interestingly, results from many of these studies have been contradictory. For instance, a study examining the effects of Tsc2 haploinsufficiency (Tsc2^+/−) on protein synthesis unexpectedly found reduced total protein levels in several mouse brain regions⁸⁰. Consistent with these findings, Auerbach and colleagues showed that increased synaptic mTOR activity suppresses protein synthesis required for long-term depression (LTD) in Tsc2^+/− mouse hippocampal slices⁸¹. On the other hand, Nie et al. performed Translating Ribosome Affinity Purification (TRAP) assay to examine ribosome-bound mRNA in Tsc2 knockdown (Tsc2-KD) rat hippocampal neuronal cultures and reported increased de novo protein synthesis⁷⁹. One prevailing hypothesis is that mTOR can up-regulate or down-regulate protein synthesis in an mRNA-specific manner. Evidence supporting this idea has been provided by several studies showing that translation of proteins is indeed differentially regulated upon loss of TSC1/2 function^79,81–83. For instance, Tsc2 deficient rodent hippocampal neurons exhibit increased expression of stress and activity-dependent proteins while some synaptic proteins are repressed^79,81. In addition, TSC2-deficient Purkinje cells derived from TSC patient iPSCs also exhibit decreased expression of proteins, particularly protein targets of the fragile X messenger ribonucleoprotein (FMRP)⁸³. These observations suggest that the TSC/mTOR axis can modulate protein synthesis bidirectionally. However, precisely how the TSC/mTOR axis mediates both the upregulation and downregulation of specific proteins remains unclear.

3.2. TSC/mTOR axis in local axonal translation

The existence of mTOR-dependent translation in neuronal somata and dendritic spines is well established¹⁹. More recent studies also implicate the mTOR kinase in local protein synthesis in axons of both the peripheral (PNS) and central nervous system (CNS)^45,84. Specifically, Terenzio et al. found that injured axons in the PNS accumulate and translate mTOR mRNA at injured sites and showed that proteins important for axonal recovery and neuron survival are locally synthesized in an mTOR-dependent manner⁸⁴. In the CNS, Poulopoulos et al. showed that mTOR accumulates in dense local foci within the growing axons of mouse cortical projection neurons, which they termed “mTOR outposts”⁴⁵ (Fig. 2). Upon closer examination, they found that these foci are highly enriched in La-related protein 1 (Larp1) which mediates mTORC1-dependent TOP mRNA translation⁸⁵, as well as TOP mRNA transcripts⁴⁵. So far, there is no direct evidence implicating TSC1/2 in these processes; however, Tsc1 is enriched in “mTOR outposts”⁴⁵ (Fig. 2). Taken together, their data strongly suggests the presence of mTORC1-dependent local translation machinery that is potentially regulated by TSC1/2.

In addition to Tsc1, the presence of Tsc2 protein has also been reported in axons of rat hippocampal cultured neurons⁴⁴. Although axonal localization of the Tbc1d7 protein has not been reported yet, the presence of mTOR, Tsc1, and Tsc2 proteins in axons and growth cones strongly suggests that the TSC/mTOR signaling axis might be well positioned to regulate the functional and structural properties of axons. Furthermore, the presence of these mTOR-dense foci could be the key to understanding how neurons sense and respond to stimuli in a precise spatiotemporal manner, although future studies are warranted to shed light on the content, location, function, and dynamics of these foci.

4.1. TSC1/2 in axon formation/specification

Neuronal polarization occurs during early development when neurons extend multiple equivalent nascent neurites, and eventually, one of them becomes an axon, while the rest become dendrites⁹⁵ (Fig. 3). The process of axon formation and specification consists of multiple steps and requires the participation and coordination of several signaling molecules and cascades^{22,35,41,44,91–94,96}.

Proper function of TSC1/2 is critical for normal neuronal polarization and axon formation as shown by several in vitro and in vivo studies^22,35,44. For example, Choi et al. showed that downregulation of TSC1/2 activity leads to the formation and maintenance of multiple neuronal axons⁴⁴ (Fig. 3). The authors discovered that TSC1/2 negatively regulates the synthesis of Synapses of the Amphids-Defective (SAD) kinases in rat hippocampal and cortical neurons as well as in cortical tubers from a TSC patient⁴⁴. Interestingly, suppression of SAD kinase expression only partially rescued the neuronal polarization abnormalities in TSC1/2 deficient neurons, suggesting that other mechanisms also contribute to this phenotype⁴⁴. Conversely, rapamycin treatment or overexpression of Tsc1 or Tsc2 proteins in hippocampal neurons in vitro led to a suppression of axon formation⁴⁴, providing further supporting evidence that proper function of the TSC/mTORC1 axis is required in these processes (Fig. 3). Another study by Zhang et al. reported that impaired peroxisomal Tsc2 localization, which is required for repression of mTORC1 and induction of autophagy in response to reactive oxygen species (ROS), resulted in deficits in neuronal morphology and polarity in rodent hippocampal cultures. With regards to axonal specification, the authors observed a reduction in the proportion of neurons without axons and an increase in the proportion of neurons with multiple axons⁴¹. In agreement with previous studies, they also showed that Tsc1/2 overexpression in hippocampal neurons increased the percentage of neurons with no axons and decreased the percentage of neurons with multiple axons⁴¹ (Fig. 3).

Interestingly, loss of Tbc1d7 also leads to abnormal neuronal axonal growth and polarization³⁵. Similar to phenotypes observed upon loss of Tsc1 or Tsc2, there was a greater proportion of Tbc1d7-null primary mouse hippocampal neurons with reduced axon length and impaired single-axon specification compared to wild type³⁵. Rapamycin treatment was sufficient to restore these phenotypes, suggesting that they are mTORC1-dependent³⁵. Taken together, these findings indicate that proper TSC1/2 activation is necessary for normal axonal formation and specification. In addition, these studies suggest that TSC1/2 functions, including control of axon specification, may be compartmentalized via differential sub-cellular localization of the complex.

The role of TSC1/2 in axonal formation and specification has been mainly explored in vitro in primary hippocampal and cortical cultures as in vivo experiments are technically challenging, particularly for neuronal populations with dense axonal projections. However, a study utilizing protein-retention expansion microscopy (ProExM) and light sheet fluorescence imaging revealed that loss of Tsc1 from mouse dopaminergic (DA) neurons led to an increase in the total axon volume of the striatal DA projections⁹⁷. This phenotype was attributed to an increase in the radius of individual axon segments as well as an increase in the axonal density⁹⁷. While there was no direct evidence, the increase in axonal density could be caused by impaired axonal specification and multiple axon formation as seen upon downregulation of TSC1/2 function.

Several studies have identified mTORC1 downstream effectors that affect neuronal polarity and axon formation, such as 4E-BP and p70S6K⁹⁸. mTORC1 itself also regulates the synthesis of several proteins involved in neuronal polarity, including the small GTPase Rap1B^93,99, SAD kinases⁴⁴, Collapsin Response Mediator Protein-2 (CRMP2), Tau¹⁰⁰ as well as the cytoskeletal regulator palladin⁹². A few studies have also attempted to elucidate the molecular mechanisms that regulate TSC1/2 during axon specification. One study provided a link between the neuronal Huntingtin-associated protein 1 (Hap1) and Tsc1⁹¹. This study showed that Hap1 is a binding partner of Tsc1 that triggers its downregulation and, consequently, the upregulation of mTORC1 signaling. Hap1 knockdown resulted in impaired axon specification and positioning of pyramidal neurons in vivo⁹¹. However, our understanding of the signaling cascade and the regulators of TSC1/2 is still very limited. Further studies are warranted to elucidate regulators and effectors of TSC1/2 and explain how these proteins coordinate to promote axonal growth and specification.

4.2. TSC1/2 in axon guidance

During nervous system development, extending axons traverse a complex environment to reach their appropriate destinations¹⁰¹. Located at the very tip of every growing axon, the growth cone navigates by following attracting or repulsive cues that eventually guide the axon to its target, enabling the formation of precise neuronal networks¹⁰¹. Early studies of the Drosophila visual system showed that loss of Tsc1 or Tsc2 from photoreceptors caused aberrant axon targeting in the CNS, and the severity of this phenotype was correlated with the levels of TOR activity^102,103. These observations uncovered the importance of the TSC/TOR axis in axon guidance.

The projection of retinal ganglion cells (RGCs) is one of the most well-characterized axon pathways in the CNS. RGC axons extend from the eye to several CNS locations, including the dorsal lateral geniculate nucleus (dLGN), by following guidance cues such as ephrins^104,105. Interestingly, RGCs in a constitutive Tsc2^+/− mouse model exhibited abnormal retinogeniculate projections¹⁰⁶. Specifically, Tsc2 haploinsufficiency caused abnormal growth cone collapse in response to ephrin treatment and led to the formation of aberrant projections in the dorsal LGN¹⁰⁶. The authors proposed that this phenotype was mediated via mTORC1 hyperactivity as they showed that ephrin receptors promote growth cone collapse via TSC-mediated mTORC1 suppression¹⁰⁶. In contrast to these results, another study showed that enhanced axonal outgrowth and impaired responses to ephrin observed in TSC2^+/− neurons were both mTOR-independent phenotypes¹⁰⁷. Specifically, forebrain cortical neurons differentiated from iPSCs derived from TSC2^+/− patients exhibited impaired axon growth and sensitivity to inhibitory axon guidance cues, as well as diminished Ras homolog family member A (RHOA) activity¹⁰⁷. The authors found that TSC2^+/− neurons had normal mTOR signaling activity, and acute treatment with mTOR inhibitors had little effect on the observed phenotypes¹⁰⁷. Therefore, they proposed that impaired axon growth and sensitivity to inhibitory cues is mTOR-independent¹⁰⁷. In addition, the authors showed that they could rescue the phenotypes via activation of the RHOA pathway¹⁰⁷, which localizes in developing axons and growth cones and regulates actin cytoskeleton and Semaphorin-3A-mediated growth cone collapse¹⁰⁸. Interestingly, these two studies may provide evidence for cell-type or species-dependent TSC1/2 mechanisms. Indeed, TSC1 was previously shown to directly interact with the ezrin-radixin-moesin (ERM) family of actin-binding proteins and activate the small GTP-binding protein RHO in primary human endothelia cells¹⁰⁹. It will be important to explore these signaling cascades in different experimental models and elucidate the downstream targets of the TSC1/2 that control axon guidance.

4.3. TSC1/2 in axon regeneration

The failure of damaged CNS axons to regenerate and form functional connections is one of the main causes of permanent disabilities. In contrast, PNS axons have a greater capacity to regenerate upon injury, enabling recovery of motor and sensory functions¹¹⁰. The molecular mechanisms that regulate the regeneration capacities of the CNS and PNS include multiple cell autonomous and non-cell autonomous factors that contribute to these processes¹¹¹.

The TSC/mTOR signaling axis has been strongly implicated in the axonal regenerative ability of the CNS¹¹²; however, the underlying mechanisms have yet to be defined. Notably, levels of mTOR activity during CNS axonal development diminish over time, and mTOR signaling is further downregulated upon axon injury¹¹². These observations suggest that downregulation of mTOR activity could potentially exert an inhibitory influence on axon regeneration. In fact, robust axon regeneration was observed in mice lacking either phosphatase and tensin homolog (Pten) or Tsc1 proteins after optic nerve injury¹¹². Importantly, loss of Pten exerted more robust regeneration capabilities than loss of Tsc1, suggesting that while mTORC1 activity indeed regulates axon regeneration in the CNS, there are additional molecular players contributing to this process¹¹². The regenerative capabilities of mTOR modulation are not limited to the optic nerve, as loss of Pten also enables corticospinal neuron sprouting and outgrowth upon spinal cord injury¹¹³. While mTOR appears to be a key player in axon regeneration, there is also evidence supporting that TSC1/2 can affect this process in an mTOR-independent manner. Specifically, in HEK293 cells, TSC1/2/Rheb has been shown to inhibit B-Raf kinase¹¹⁴, a critical component of the MAPK signaling pathway¹¹⁵. B-Raf activation was found to promote axon development and regeneration both in mouse CNS and in PNS¹¹⁶. It will be important to understand how and when TSC1/2 orchestrates mTOR and other proteins to ensure proper axon development and regeneration.

5.1. Myelination in the CNS: TSC1/2 effect on oligodendrocytes

Most CNS myelination occurs postnatally and continues throughout adulthood in a spatially and temporally regulated manner as neuronal circuits mature. During myelination, populations of oligodendrocyte progenitor cells (OPCs) transition into mature myelinating cells by undergoing extensive cytoskeletal changes involving large-scale synthesis and trafficking of myelin-related lipids and proteins^125–128 (Fig. 4A). Expression pattern analysis of Tsc1 and Tsc2 revealed that the levels of these proteins gradually decrease during OPC maturation, but they remain present in all OL lineage cells¹²⁹ (Fig. 4A). This observation suggests that there is a temporal regulation of TSC1/2 and its downstream signaling, including mTORC1, which is important for the proper development and maturation of myelinating cells.

Several studies using conditional knockout (KO) mouse models have examined the impact of loss of either Tsc1 or Tsc2 genes in OLs at various developmental stages and within different CNS regions^129–132. Conditional deletion of Tsc2 in OPCs from mice expressing Cre recombinase driven by the Olig2 promoter (Olig2-Cre;Tsc2-KO) showed the expected increase in mTORC1 and decrease in Akt/mTORC2 activity in the cerebral cortex, increased gliosis, and significant reduction in CNS myelination¹³⁰. The authors suggest that loss of myelination may be due to a cell fate shift from OL to astrocyte lineage¹³⁰. Ablation of Tsc1 from a late subpopulation of OPCs in the cyclic nucleotide phosphodiesterase (CNP)-Cre mouse line¹³³ also showed strong hypomyelination accompanied by impairment to the sterol regulatory element-binding proteins (SREBPs) pathway, a master regulator of lipogenic and sterolgenic gene transcription^131,134. Further studies in the CNP-Cre;Tsc1-KO model report impaired OL differentiation and abnormal organization of myelin across the axon, which were attributed to downregulation of the RNA binding proteins Quakin and Neurofascin¹³². Another study showed that loss of Tsc1 from an OL lineage-expressing Olig1 (Olig1-Cre mouse line) caused an increase in OPC proliferation during early developmental stages; however, during differentiation stages, Tsc1-KO OPCs exhibited elevated endoplasmic reticulum (ER) stress responses, activation of apoptotic pathways and cell death leading to a severe hypomyelination phenotype in vivo¹²⁹. Treatment with rapamycin in animal models with impaired Tsc1/2 function in myelinating cells resulted in a partial but significant improvement in myelination suggesting that mTORC1 plays a crucial role in the development of the hypomyelination phenotype^129–132; however, the contribution of other pathways cannot be fully excluded.

While the role of mTORC1 signaling in regulating myelination has been established, the role of mTORC2 is less understood; loss of mTORC2 signaling appears to have minimal effect on developmental myelination yet, later in development, it can cause an increase in the number of mature OLs¹³⁵.

Interestingly, the mTOR pathway appears to have distinct effects on brain myelination depending on the developmental stage. Although loss of mTORC1 delayed OL maturation in the corpus callosum (CC) of 2-month-old CNP-Cre;Rptor-KO mice, there was no significant change in myelination¹³⁵. In addition, electron microscopy of 2-month-old CNP-Cre;mTOR-KO mice show structurally normal myelination in the brain¹³⁶. However, analysis of the CC in 3-month-old mice showed significant reduction of myelin basic protein (MBP) and myelin proteolipid protein (PLP) expression, fewer myelinated axons, and thinner myelin sheaths¹³⁶. In addition, inducible Plp-CreERT;mTOR-KO adult mice revealed that loss of mTOR from mature OLs impairs myelination integrity¹³⁶. These studies, taken together, highlight the importance of mTOR both in developmental myelination and in myelin integrity in the adult brain.

Several studies also provide evidence that OLs from different CNS regions are transcriptionally and functionally distinct^136,137. Specifically, single-cell sequencing analysis of late-stage OPCs from mouse brains and spinal cords revealed that brain OPCs have reduced capacity for cholesterol biosynthesis in comparison to spinal cord OPCs that is potentially compensated for by higher receptor-mediated cholesterol uptake¹³⁶. In addition to the cholesterol biosynthesis pathway that is known to be regulated by mTORC1¹³⁸, cytoskeletal organization-related pathways that are known to be regulated by mTORC2 were differentially expressed between OPCs from different regions of the nervous system. This could explain the differential sensitivity of the different OL populations to mTOR complex-specific perturbations.

Several different mechanisms have been implicated in hypomyelination phenotypes due to TSC1/2 loss of function in OL cells, including defective lipogenesis, impaired OPC differentiation, and apoptosis (Fig. 4A). These different findings could be attributed to the use of different Cre lines that cause TSC/mTOR signaling perturbations in different OL subtypes and lineages. Further studies are warranted to explore how the TSC/mTOR signaling pathway affects myelination during different developmental stages and across OL differentiation.

5.2. Myelination in the CNS: non-cell autonomous effects of TSC loss in neurons

Surprisingly, mouse models with neuron-specific Tsc1 deletion also exhibit severe hypomyelination^22,139, suggesting that neurons can regulate myelination in a non-cell autonomous manner. For example, loss of Tsc1 from mouse neurons leads to decreased levels of MBP, increased populations of OPCs, and reduced populations of mature OLs^22,139. Ercan et al. found that Tsc1-KO mouse neurons and neurons derived from TSC patient iPSCs also exhibit elevated secretion of connective tissue growth factor (CTGF)¹³⁹. The authors showed that the addition of CTGF to OPC cultures in vitro, or enhanced expression of CTGF in wild type mouse brains in vivo, was sufficient to profoundly reduce the numbers of MBP+ OLs¹³⁹. These findings suggest that CTGF negatively regulates myelination, thus providing a mechanism by which neurons can influence myelinating cells (Fig. 4A). Finally, rapamycin treatment in vivo restored CTGF expression and improved myelination, suggesting that these phenotypes are mTOR dependent¹³⁹. In agreement with this idea, a study by Zou et al. showed that loss of Rheb1 from neuronal progenitor cells (NPCs) using the Nestin-Cre mouse line led to downregulation of mTORC1 signaling, upregulation of mTORC2, and profound myelination deficits¹⁴⁰, further implicating mTOR signaling in myelination.

Several studies also report that axon diameter and structure are impaired in different TSC models^8,22,97,141. For example, SynI-Cre;Tsc1-KO mice exhibit axonal outgrowth that abnormally persists in later developmental stages²². In addition, deletion of Tsc1 from mouse dopaminergic neurons increased individual axon radius and sprouting⁹⁷. Finally, tuber and peri-tuber samples from TSC patients exhibit increased axonal connectivity, as well as thicker axons with longer and more elaborate branches⁸. Hence, it would also be important to explore whether and how TSC-related changes in axonal structure impact myelination and contribute to these deficits.

Together, these studies support the notion that changes in TSC/mTOR signaling can affect myelination via cell autonomous or non-cell autonomous mechanisms and highlight the importance of proper neuron-glia communication. Further studies are necessary to elucidate the contribution of glia and neurons in the TSC hypomyelination phenotypes as well as the exact mechanisms that underlie them.

5.3. Myelination in the PNS: TSC1/2 protein complex effect on Schwann Cells

Like OLs, Schwann cells (SCs), the main glia of the PNS, also require the coordination of several regulatory molecules and signaling cascades to myelinate axons properly¹⁴² (Fig. 4B). Figlia et al. used the Dhh-Cre;Tsc1-KO mouse model to show that SC-specific deletion of Tsc1 leads to significant delay of myelination onset due to arrest of SCs at the promyelinating stage¹⁴³. The authors further discovered that the transcriptional factor Krox20, a key driver of PNS myelination, is negatively regulated by mTORC1 activity¹⁴³. Rapamycin treatment was sufficient to restore Krox20 expression and improve myelination, suggesting a potential molecular link between mTOR and myelination in the PNS¹⁴³. In addition, Jiang et al. also showed that loss of Tsc1 from SC progenitors results in impaired myelination due to inhibited differentiation and increased proliferation of progenitor SCs as well as delayed radial sorting¹⁴⁴. The authors, in agreement with previous studies, also found that Krox20 was downregulated and reported altered expressions of polo-like kinases (PLK)¹⁴⁴, which are involved in cell proliferation and mitosis¹⁴⁴. Similar hypomyelination phenotypes, as well as impaired SC differentiation and proliferation, were observed by Beirowski et al., who studied loss of Tsc2 from SC lineage using the P0-Cre;Tsc2-KO mouse line¹⁴⁵. Together, these studies indicate that TSC1/2 activity tightly regulates SC differentiation, and that suppression of mTORC1 signaling is necessary to allow the onset of proper myelination in the PNS (Fig. 4B).

Interestingly, the effects of deregulation of the TSC/mTOR axis on PNS myelination strongly depend on the developmental stage of SCs. Loss of Tsc1/2 function in SCs early in development and prior to the onset of myelination leads to hypomyelination¹⁴³. On the other hand, loss of Tsc1/2 function in adult SCs after myelination onset results in radial hypermyelination (i.e., increased myelin thickness without visible structural abnormalities)¹⁴³. These observations uncover a dual role for the TSC/mTOR axis that depends upon the differentiation status of SCs. Notably, while TSC animal models exhibit mTOR-related impaired myelination in both the CNS and PNS, myelin-related PNS abnormalities have not yet been reported in TSC patients. If human PNS myelination is not dependent on mTOR activity in SCs, understanding the molecular mechanisms driving these species-specific differences would be important.

5.4. mTOR dependent and independent mechanisms in myelination

As previously discussed, loss of TSC1/2 function in rodent and human models results in severe hypomyelination that can be driven by both cell autonomous and non-autonomous mechanisms. Rapamycin treatment significantly improves these phenotypes, suggesting that myelin abnormalities are driven primarily by mTOR signaling hyperactivity. However, these findings have introduced an interesting paradox to the field given that mTOR is known to be a positive regulator of not only cell differentiation, protein, and lipid biosynthesis, but also myelination (extensively reviewed elsewhere)^63,124.

In fact, studies examining direct inhibition of mTORC1 and mTORC2 signaling via deletion of their essential components, Raptor^146,147 and Rictor⁶¹, respectively, revealed that in the CNS, mTORC1, and to a lesser extent mTORC2, both positively regulate myelination^135,148,149. Specifically, mTOR signaling is required for OL cytoskeletal organization and differentiation, as well as initiation and completion of CNS myelination^135,148,149. In the PNS, mTORC1 controls lipid biosynthesis via SREBPs, and its downregulation delays and reduces myelination¹⁵⁰. Ablation of mTOR using the CNP-Cre line also revealed impaired postnatal growth of myelinating SCs¹⁵¹. Interestingly, while several studies have shown that the two mTOR complexes regulate one another⁵³, they seem to differentially affect myelination in a region-specific manner. For instance, it was shown that mTORC2 downregulation affects myelination primarily within the brain but mildly affects the spinal cord¹⁵². In contrast, mTORC1 is a positive regulator of OL differentiation in the spinal cord, and its suppression leads to significant demyelination^135,153.

TSC-related phenotypes, including hypomyelination, have been typically attributed to mTORC1^55,141. However, some of the findings discussed here raise the possibility that TSC1/2 could also exert mTOR-independent functions and impact myelination. For example, Lebrun-Julien et al. found that downregulation of the SREBP pathway and impaired lipogenesis is a common phenotype among CNP-Cre;Tsc1-KO and CNP-Cre;Rptor-KO mice¹³¹. Given that mTORC1 is a positive regulator of SREBPs^154,155, this phenotype was expected in the Rptor-KO mice but unexpected in the Tsc1-KO mice. It has been previously shown that in Tsc1-null hepatic cells the SREBP pathway could be modulated in an mTORC1-independent and Akt-dependent manner¹⁵⁶, suggesting that Akt might be an alternative mechanism via which TSC1/2 can affect myelination.

Akt has been identified as a key player in myelination, and while some of its functions are attributed to mTOR signaling, others appear to be independent¹⁵⁰. Overexpression of Akt in the CNS of mice was found to increase myelin production per OL, causing hypermyelination without affecting the numbers of OPCs or OLs^157,158. This hypermyelination phenotype was attributed to mTORC1 as it was found to be upregulated, and rapamycin treatment was sufficient to improve it. However, chronic rapamycin treatment can also suppress mTORC2/Akt signaling in the CNS¹⁵⁹, suggesting that the observed rescue could be mTORC1-independent. In addition, a study comparing the loss of Tsc1 and Pten from adult SCs reported elevated mTORC1 activity and hypermyelination but also observed subtle differences in the myelination phenotypes¹⁴³. These include differences in radial hypermyelination, suggesting that an PI3K-Akt-dependent/mTORC1-independent mechanisms may be involved in SC membrane wrapping¹⁴³. Pten is also known to regulate OL development via an Akt-dependent, mTOR-independent mechanism¹⁶⁰. Since loss of TSC1/2 activity suppresses mTORC2/Akt⁵³, this may be an alternative pathway via which TSC1/2 controls myelination independent of mTORC1. Further studies are necessary to elucidate what mTOR-independent mechanisms are involved and how much these contribute to the myelination deficits.

6.1. TSC1/2 in CNS remyelination

To study the effects of TSC1/2 on remyelination in different OL lineages, McLane et al. used an inducible Cre-lox system to explore how loss of Tsc1 affects remyelination after focal lysolecithin-mediated lesions¹⁶⁷. Notably, while loss of Tsc1/2 function in OPCs during early developmental stages causes hypomyelination^129–132, McLane et al. discovered that loss of Tsc1 from nerve/glial-antigen 2 (NG2)+ adult OPCs accelerated myelin protein expression and remyelination, resulting in increased myelin thickness¹⁶⁷ (Fig. 5B). Interestingly this increase in myelination was not accompanied by changes in the number of OPCs or OLs¹⁶⁷, indicating that loss of Tsc1 in adult OPCs does not impact OL differentiation or proliferation. This finding agrees with previous studies showing that inhibition of mTOR via rapamycin impairs OL metabolic function and significantly delays remyelination in the cuprizone demyelination mouse model^168,169. In contrast, when the authors examined loss of Tsc1 in PLP+ differentiating OLs, they found that the speed and thickness of myelination were transiently impaired¹⁶⁷ (Fig. 5B). One potential explanation McLane et al. provide for the differential impact of TSC in remyelination relies on the differential effects of mTORC1 and mTORC2 in OL development.

6.2. TSC1/2 in PNS remyelination

To explore the role of TSC1/2 in PNS remyelination, Figlia et al. performed nerve crush injuries in mice with specific loss of Tsc1 in SCs¹⁴³. The authors reported that while no significant differences in demyelination were observed following injury, SC redifferentiation and remyelination in Tsc1 deficient mice were significantly delayed¹⁴³ (Fig. 5D). In addition, upon completion of the remyelination process, Tsc1-KO mice exhibited a significant reduction in myelinated fibers¹⁴³. In agreement with the distinct roles of the TSC/mTOR axis across development, the authors found that loss of Tsc1 from SCs that had just differentiated into myelinating SCs augmented myelination using a tamoxifen-inducible line¹⁴³ (Fig. 5D). Interestingly, while loss of Tsc1 and consequent mTORC1 hyperactivation did not affect dedifferentiation¹⁴³ in one study, another study showed that suppression of mTORC1 activity reduced myelin clearance and impaired SC dedifferentiation¹⁷⁰. One potential explanation the authors suggest is that mTORC1 activity is maximal during normal dedifferentiation, and further increase of mTORC1 signaling does not exert any effects. However, during redifferentiation and remyelination mTORC1 signaling declines, hence its upregulation is detrimental for proper myelination^143,170. Further studies are necessary to explore how TSC/mTOR activity fluctuates during remyelination steps and whether specific TSC1/2-related molecular mechanisms could be targeted to enhance remyelination of the PNS upon injury.