Oligodendrocytes (OLs), the myelin-forming CNS glia, are highly vulnerable to cellular stresses, and a severe myelin loss underlies numerous CNS disorders. Expedited OL regeneration may prevent further axonal damage and facilitate functional CNS repair. Although adult OL progenitors (OPCs) are the primary players for OL regeneration, targetable OPC-specific intracellular signaling mechanisms for facilitated OL regeneration remain elusive. Here, we report that OPC-targeted PTEN inactivation in the mouse, in contrast to OL-specific manipulations, markedly promotes OL differentiation and regeneration in the mature CNS. Unexpectedly, an additional deletion of mTOR did not reverse the enhanced OL development from PTEN-deficient OPCs. Instead, ablation of GSK3β, another downstream signaling molecule that is negatively regulated by PTEN-Akt, enhanced OL development. Our results suggest that PTEN persistently suppresses OL development in an mTOR-independent manner, and at least in part, via controlling GSK3β activity. OPC-targeted PTEN-GSK3β inactivation may benefit facilitated OL regeneration and myelin repair.https://doi.org/10.7554/eLife.32021.001
In the mammalian central nervous system (CNS), NG2 and PDGFRα (platelet-derived growth factor receptor α)-expressing OPCs (also known as NG2+ cells) are abundant throughout life (Dimou and Gallo, 2015; Nishiyama et al., 2009). The lineage progression from OPC to OL is a multi-step process that involves gradual changes in cell morphology and properties, from OPCs to pre-myelinating OLs (pre-OLs) to fully mature OLs, culminating in the formation of lipid-rich myelin around axons (myelinogenesis) (Nave and Werner, 2014). In the adult CNS, OPCs continue to serve as the primary reservoir for homeostatic maintenance of OLs by generating new myelinating OLs in the event of OL loss (Chamberlain et al., 2016; Young et al., 2013), while they proliferate and sustain their own population size (Hughes et al., 2013).
Although a number of intracellular signaling pathways are known to regulate OL development and myelination during early development (Gaesser and Fyffe-Maricich, 2016), it is not clear whether the same signaling mechanisms equally regulate continuing OL generation in the adult CNS, as well as OL regeneration after demyelination (Fancy et al., 2011). Indeed, genetic ablation of proposed OL development regulators often resulted in only transient delay in early myelination (He et al., 2017; Huang et al., 2013). Moreover, in contrast to a plethora of molecules whose loss-of-function results in myelin defects, there are far fewer examples of significantly improved OL generation and regeneration in the mature CNS.
It is conceivable that there are multiple cell-stage-specific signaling mechanisms, each of which may regulate a distinct aspect of the OL development and myelination, such as OPC proliferation, a permanent exit from the OPC cell cycle, morphological change, axon recognition, and, finally, plasma membrane outgrowth and tight myelin formation. Accordingly, OPC-intrinsic signaling that regulates OPC-to-OL transition should be considered as a separate step from the OL-intrinsic mechanisms that drive myelin growth and maintenance.
The mammalian target of rapamycin (mTOR) signaling is known to be necessary for proper CNS myelination in early life (Figlia et al., 2018; Lebrun-Julien et al., 2014; Wahl et al., 2014; Wood et al., 2013), but the therapeutic potential of mTOR signaling manipulation for improved OL regeneration is not clear (Jiang et al., 2016; Lebrun-Julien et al., 2014; McLane et al., 2017). As an effort to identify OPC-intrinsic signaling mechanisms that promote OL generation in the mature CNS, we first determined whether the OPC-specific increase in mTOR complex 1 (mTORC1) activity enhances OL development, by OPC-specific genetic deletion of the tuberous sclerosis complex 1 (TSC1) or phosphatase and tensin homolog (PTEN), the two negative regulators of mTORC1 activity (see Figure 1A). Here, we report that OPC-targeted PTEN genetic ablation markedly enhanced OPC proliferation and OL differentiation, regardless of the examined CNS areas and age windows. OPC-targeted PTEN inactivation also remarkably facilitates OL regeneration and promotes remyelination after toxin-induced demyelination. However, such positive outcomes were not observed for OL-specific PTEN ablation. Moreover, in contrast to PTEN, OPC-specific TSC1 inactivation impaired OL development/survival.
The conflicting outcomes of genetic inactivations of PTEN and TSC1 suggest involvement of mTORC1-independent PTEN-downstream signaling mechanisms. Supporting this, our Pten and Mtor double cKO revealed that mTORC1 signaling is dispensable for the enhanced OL generation from the PTEN-deficient OPCs. Finally, OPC-specific ablation of glycogen synthase kinase 3β (GSK3β), another PTEN-Akt downstream target, promoted OL differentiation at a level comparable to that of PTEN-ablation.
We interpret our findings to indicate that PTEN persistently and negatively regulates the OPC-to-OL transition in an mTOR-independent manner, likely via GSK3β activity. This suggests that the inhibition of the PTEN-Akt-GSK3β pathway may efficiently promote remyelination in the adult CNS.
Akt-mTORC1 signaling regulates early OL development and myelination (Figlia et al., 2018; Lebrun-Julien et al., 2014; Wahl et al., 2014; Wood et al., 2013). Because the mTORC1 activity is under inhibitory control of the TSC1/2 complex (Tee et al., 2002) (See Figure 1A), it could be expected that Tsc1 ablation would increase mTORC1 activities, and enhance oligodendrogenesis. However, contrary to this idea, Tsc1 cKO with Cnp-Cre mice resulted in impaired myelination (Lebrun-Julien et al., 2014). The Cnp-Cre is active in mature OLs (Lappe-Siefke et al., 2003), but it impacts only a subset of late (i.e. GPR17-expressing) OPCs during early development (Tognatta et al., 2017). Therefore, we reasoned that the hypomyelination observed in Cnp-Cre-based Tsc1 cKO may be attributed to long-term OL-specific defects, such as myelin maintenance failure, rather than to the OPC-to-OL transition.
To test the possibility that the OPC-specific mTORC1 activity increase promotes OL development in vivo, we crossed Pdgfra-CreERTM mice (tamoxifen-inducible OPC-specific Cre mice) (Kang et al., 2010) with Tsc1f/f and Rosa26-EYFP (R26-EYFP) Cre reporter mice. 4-Hydroxytamoxifen (4HT) was administered into Pdgfra-CreER; R26-EYFP; Tsc1f/f mice starting at P20 (1 mg per i.p. injection, a total of three injections for 1.5 days) and their age-matched controls (Pdgfra-CreER; R26-EYFP) (Figure 1B). This 4HT injection protocol allows EYFP-labeling of more than 75% of OPCs in both the cortex (CTX) and corpus callosum (CC) of the control mice in 4 days (P20 +4) (data not shown). Three weeks after the first 4HT injection (P20 +21), there was a significant increase in the percentage of phosphorylated S6 ribosomal protein (pS6) immunopositive cells among all callosal EYFP+ cells in Pdgfra-CreER; R26-EYFP; Tsc1f/f mice (p=0.042) (Figure 1C,D), indicating that mTORC1 activity is enhanced (Hara et al., 1998) as a result of effective TSC1 inactivation.
However, to our surprise, the number of EYFP+ cells was remarkably reduced in Tsc1 cKO mice, particularly in the CC (Figure 1E,F). Co-immunostaining with antibodies against OL lineage stage markers (i.e. NG2 and CC1) (Figure 1E) revealed that the number of newly generated OLs (EYFP+CC1+) was severely reduced, both in the CC (p=0.016) and CTX (p=0.0007) of the cKO mice (Figure 1G). In contrast, the numbers of EYFP+ pre-myelinating OLs (pre-OLs, NG2- CC1-) and EYFP+ OPCs (NG2+) were not significantly affected, except for cortical pre-OLs that were increased in the Tsc1 cKO (Figure 1H,I). Consequently, despite no or only a small increase in labeled OPCs and pre-OLs, the composition of EYFP-labeled cells changed to a state where the proportion of OPCs and pre-OLs became significantly larger, whereas that of mature OLs was markedly reduced in both the CC and CTX (Figure 1J). The altered composition of EYFP-labeled cells excludes the possibility that the marked decrease of EYFP+ OLs was not caused by inefficient Cre-loxP recombination in Tsc1 cKO mice (P20 +21). The same OPC-targeted Tsc1 cKO mice also exhibited impairment of new OL development in the spinal cord (SC) (Figure 1—figure supplement 1).
Our examination on the Ki67 (a marker of a proliferating cell) expression pattern in the brain (Figure 1K) revealed that the OPC proliferation was not impaired in these Tsc1 cKO mice (Figure 1L). The percentage of Ki67+ cells among cortical EYFP+ OPCs was even higher in the cKO mice than in control (Figure 1L). Nonetheless, there was a slight reduction in total OPCs in the CTX of the cKO mice (Figure 1M). These observations indicate that the marked decrease of new OLs in the cKO mice is not due to OPC proliferation defects. Consistent with the impaired OL accumulation, the levels of MBP expression were also reduced in the CTX of Tsc1 cKO mice (P20 +21) (Figure 1N,O).
These results raise two possibilities that are not mutually exclusive: Pre-OL-to-OL lineage progression is arrested, and/or newly born OLs undergo premature cell death. Although our cleaved caspase-3 staining did not reveal any increase in cell death in the cKO mice brain (data not shown), it is possible that the observation time point of cleaved caspase-3 (3 weeks after Tsc1 deletion) is too late to detect the cell death events. The different degree of OL impairment in the CC and CTX may be related to the numbers of new OLs in the two brain areas for the examined age window (from P20 to P41). A more actively OL-generating CNS region may have been more severely affected by a premature OL loss and/or maturation arrest. Indeed, the CC is known to be more active in OL development than CTX in all examined ages in earlier studies (Kang et al., 2010; Rivers et al., 2008; Young et al., 2013).
Along with others’ results on OL lineage cKO of Tsc1 or Tsc2 (Carson et al., 2015; Jiang et al., 2016; Lebrun-Julien et al., 2014), our results further support the idea that a complete inactivation of TSC1 (or consequent mTORC1 hyperactivation) is detrimental to OLs, and thus would not benefit OL development or remyelination. However, the situation is not entirely clear, because another, recently published study suggests that Tsc1-deleted OPCs (with Cspg4-CreER) leads to increased myelin thickness after demyelination (McLane et al., 2017).
Next, we targeted Pten, another upstream negative regulator of the Akt-mTORC1 signaling (See Figure 1A), in OPCs. Pten had been targeted in OL lineage cells in earlier studies, in which gene deletion was achieved using either Olig2-Cre (Harrington et al., 2010; Maire et al., 2014) or OL-specific Cre lines (i.e. Cnp-Cre and Plp1-CreER) (Goebbels et al., 2010). Those studies did not report apparent changes in OL numbers or pattern of remyelination, but abnormal myelin outgrowth. However, it is possible that the cell-specific roles of PTEN in OPC were not completely revealed, because Olig2-Cre may affect many other neural cells besides OL lineage cells (Maire et al., 2014; Masahira et al., 2006), and OL-specific Cre mice do not impact the majority of adult OPCs (Goebbels et al., 2010). To investigate OPC-specific roles of PTEN, we administered 4HT into Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice either at P20 or P45, sampling them at P41 or P75, respectively (Figure 2A), and measured the degree of new oligodendrogenesis from the EYFP-labeled OPCs for these two age windows (i.e. P20 ~P41 and P45 ~P75). Similar to Tsc1 cKO, Pten deletion increased pS6 levels in EYFP+ cells in the CC (Figure 2B,C), indicating that PTEN was inactivated, and as a consequence, mTORC1 activity was significantly enhanced. However, in contrast to Tsc1 cKO, the number of EYFP+ cells increased for most of the observed brain areas, including the CC and CTX (Figure 2D,E). The fate analysis of EYFP-labeled OPCs with co-immunostaining of cell stage markers (Figure 2D) revealed that there were significant increases in EYFP+CC1+ cells in the CTX of Pten cKO mice for both age-windows (p=0.0002 for P20 ~P41; p=0.0037 for P45 ~P75) (Figure 2F). The new OL generation was also higher in the CC of P45 +30 Pten cKO mice than the age-matched control (p=0.0006) (Figure 2F). The failure to observe a significant increase of EYFP+CC1+ cells in the CC of P20 +21 Pten cKO (Figure 2F) may be due to the high rate of OL differentiation (with a larger variation) in the CC of the developing brain (Kang et al., 2010). Thus, facilitated OL differentiation may not have been sufficiently pronounced by the relatively short-term (i.e. 3 weeks) OPC fate-tracking. In contrast to the increased new OLs, densities of EYFP-labeled OPCs were not changed by Pten cKO (Figure 2G). Consequently, the percentage of mature OLs among EYFP-labeled OL lineage cells increased in both the CC (for P45 +30) and CTX (Figure 2H). These results strongly suggest that PTEN inactivation in OPCs promotes new OL development, even in the mature brain.
In order to assess changes in cell proliferation, BrdU was also given to Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice (Figure 2A). We observed that there were marked increases in BrdU+ cells in the Pten cKO brain (P20 +21) (Figure 2I left). Confocal imaging further confirmed that most BrdU+ cells were NG2+ OPCs for both age windows (Figure 2I,J). It was also noted that the characteristic tiled distribution of OPCs (Hughes et al., 2013) was disrupted by OPC-specific Pten cKO (Figure 2K). However, despite the increase of OPC proliferation, the density of total OPCs was not altered (Figure 2L). These observations suggest that the increased OPC proliferation directly contributes to additional OL differentiation and/or that it is a homeostatic regenerative response to the OPC reduction due to enhanced OPC-to-OL conversion in Pten cKO mice.
The enhanced oligodendrogenesis was also evident in both white and gray matter (WM and GM) of the ventral SC of Pten cKO mice, for both examined age windows (Figure 2—figure supplement 1A–C). In particular, after OPC-specific Pten ablation starting at P45, new OL accumulation was enhanced by about four-fold in the SC GM in 30 days (p=0.0001) (Figure 2—figure supplement 1C). However, the densities of labeled OPCs and total OPCs were not altered (Figure 2—figure supplement 1D,E), as they were in the brain. These results further suggest that OPC-specific PTEN inactivation effectively promotes new OL development throughout the CNS.
Even though OL generation is enhanced by the OPC-targeted genetic manipulation, survival of new OLs may be contingent on the availability of extracellular growth factors or other cell-extrinsic cues. Thus, we asked whether PTEN-deficient OPCs continue to add new OLs, leading to an increase in total OLs. To quantify total OLs unambiguously, we used Mobp-EGFP BAC Tg mice in which EGFP is expressed only in mature OLs (Kang et al., 2013). Pdgfra-CreER; Mobp-EGFP; ±Ptenf/f mice received three 4HT injections (a total of 3 mg) between P14 and P18 (Figure 3—figure supplement 1A). At P30, there were increases in both EGFP+ OLs (Figure 3—figure supplement 1B,C) and Olig2+CC1+ OLs (Figure 3—figure supplement 1D,E) in the CTX of Pten cKO mice.
For a longer-term OPC fate analysis, the 4HT-administered (at P20) Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice were observed at P90 (thus P20 +70) (Figure 3A). These cKO mice exhibited even greater accumulation of EYFP+ OLs in both the CC and CTX (a seven-fold increase of OL at P20 +70 vs. a two-fold increase at P20 +21) (Figures 3B–D and and 2F). Due to the greater new OL addition, total CC1+ OLs also increased (Figure 3E). Interestingly, EYFP+NG2+ OPCs were significantly reduced in the CC (p=0.0004), but not in the CTX of Pten cKO mice (P20 +70) (Figure 3F). Due to the marked EYFP+ OL accumulation, the percentages of mature OLs among EYFP-labeled cells were increased, both in the CC and CTX, whereas those of labeled OPCs were highly reduced (Figure 3G). It is likely that the PTEN-deficient (EYFP+) OPCs in the CC may have undergone OL differentiation more rapidly than their own cell division, and thus EYFP-labeled OPCs markedly decreased at P20 +70. At the same time, PTEN-harboring EYFP– OPCs may have replaced the EYFP+ OPCs. Indeed, at P20 +70, EYFP-Olig2+NG2+ OPCs were predominant in the CC of Pten cKO mice (Figure 3H), and thus, the density of total OPCs was not altered (Figure 3I).
BrdU administration (50 mg/kg per injection, a total of 10 injections between P80 and P84) and cell proliferation analyses revealed that OPCs continuously proliferate at higher rates in the CTX of Pten cKO (Figure 3J,K), and yet without changes in total OPC number (Figure 3I). Some of the BrdU-laden OPCs appeared to have become pre-OLs or OLs (EYFP+NG2-) during the last week before sampling (Figure 3K right).
These results further confirm that PTEN regulates OPC differentiation, not only in early postnatal life but also at later adult ages. These also suggest that the enhanced OL differentiation by PTEN inactivation is not a transient event, and that the newly generated OLs from PTEN-deficient OPCs survive long. Although OPC proliferation remained promoted in the CTX, it appears that this enhanced proliferation is tightly controlled to maintain OPC homeostasis.
Although in sharp contrast to earlier studies in which no changes (Goebbels et al., 2010; Harrington et al., 2010) or a decrease (Maire et al., 2014) in OL numbers were observed with different Cre-dependent Pten deletion, our results provide an important example of how the outcomes of OPC-specific genetic manipulations can differ from those of mature OL-specific or non-cell-specific cKO when assessing cell-intrinsic regulators for adult OL development.
In order to determine whether the increased new OLs contribute to enhanced myelination in Pten cKO mice, we performed western blot analysis for myelin proteins. 4HT was injected into Ptenf/f (control) or Pdgfra-CreER; Ptenf/f mice four times between P14 and P21 (a total of 4 mg) (Figure 4A). Even though Pten might have been deleted from only subsets of young OPCs by this Cre activation protocol, levels of PLP, MOG, and CNP were significantly increased in the CTX of Pten cKO mice at P75 (Figure 4B,C). Moreover, electron microscopic (EM) analyses revealed that the g-ratios (the ratio between the inner and the outer diameters of the myelin sheath) of callosal axon fibers were significantly lowered in the Pten cKO mice (P13 +17) (p=0.038) (Figure 4D–F). More importantly, there was a marked increase in the number of myelinated axons (p=0.033) (Figure 4G), suggesting that the increased number of new OLs myelinate more axons in OPC-specific Pten cKO mice.
To better observe the growth of OL processes, Pdgfra-CreER; ±Ptenf/f mice were crossed with R26-mEGFP (mT/mG) mice, which express the membrane-bound EGFP upon Cre recombination (Muzumdar et al., 2007). In a few weeks after 4HT administration (at P25), EGFP+ cells exhibited more rapid morphological changes in the Pten cKO mice, and the EGFP+ OPC progeny cells formed more thin processes along axonal tracts (Figure 4H,I). Consequently, at P25 +20, discrete stellate-shaped OPC morphology was no longer seen in the CC of cKO mice (Figure 4H). We also performed immuno-EM with Pdgfra-CreER; R26-mEGFP; ±Ptenf/f (P20 +21) mice and anti-EGFP antibodies, in which the gold particles labeled only EGFP+ membranes, including newly formed myelin sheaths (Figure 4J). The immuno-EM analysis revealed that the percentage of newly formed myelinated axons increased by about 2.5-fold in the CC of the Pten cKO mice (p=0.0179) (Figure 4K). These results suggest that PTEN inactivation in OPCs facilitates and enhances new myelination.
In two earlier independent studies, Akt signaling was activated in OLs by using Plp1-Akt1 (DD) transgenic mice (Flores et al., 2008) or Cnp-Cre; Ptenf/f mice (Goebbels et al., 2010). Both mutant mice exhibited similar phenotypes, such as thicker myelin sheaths, but with no change in OL number and OL regeneration. However, the gene promoters utilized (i.e. Plp1 and Cnp) in those studies may have affected at most only a small fraction of OPCs, given their endogenous gene expression patterns. Thus, the phenotypic differences between our results and those earlier ones may have stemmed from different target cell populations (i.e. OPC vs. OL). Alternatively, it is also possible that different methods utilized in our OL lineage assessments caused these differences. To determine whether differences in target cell populations for PTEN-Akt manipulation account for the observed phenotypic differences, we used Mog-iCre mice (Buch et al., 2005) for Pten deletion. In Mog-iCre; R26-EYFP; ±Ptenf/f mice (P41), EYFP was expressed only in CC1+ OLs (Figure 5A), suggesting that the Cre activity is highly specific to OLs. The numbers of EYFP+ OLs and NG2+ OPCs were not changed by the Mog-iCre-based Pten cKO for both the brain and the spinal cord (Figure 5B,C). Moreover, EM analysis of R26-EYFP; Ptenf/f; ±Mog-iCre (P34) mice revealed that myelin sheaths were thickened (Figure 5D–F), but the percentage of myelinated axons did not increase in the cKO mice (Figure 5G). These results are in accordance with the earlier reports (Flores et al., 2008; Goebbels et al., 2010), and together they support the notion that the enhanced oligodendrogenesis (promoted OL generation and increase in myelinated axons) in Pdgfra-CreER; Ptenf/f mice is due to OPC-specific targeting of Pten, thus suggesting differential cell-stage-specific roles of PTEN.
Given the facilitated OL development in OPC-specific Pten cKO mice, we asked whether Pten-inactivated OPCs also enhance OL regeneration. After Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice received tamoxifen (a total of 10 injections for 5 days) from P50 to P54, 1% lysolecithin (LPC) was injected into the CC at P58 (Figure 6A). This LPC injection typically induces focal demyelination in 3 days (Figure 6B), and OL regeneration (and remyelination) occurs for the following 3 weeks. Two weeks after LPC, we observed that more EYFP+CC1+ OLs accumulated in the CC of Pten cKO mice (p=0.01) (Figure 6C–E), while EYFP-labeled OPCs were unchanged (Figure 6F). Thus, the percentage composition of EYFP-labeled cells changed, which indicates more OL generation from OPCs (Figure 6G). EM analyses also revealed that myelin thickness was increased, particularly for large-diameter axons (diameters > 1 μm, p=0.0458) in the Pten cKO mice (Figure 6H–K).
We also injected LPC into the dorsal SC WM (Figure 6—figure supplement 1A,B). To this cohort of mice, BrdU was also administered from 1 dpi (1 day after LPC injection) until mouse sampling (Figure 6—figure supplement 1A). We observed that more EYFP+CC1+ OLs accumulated in the lesions of Pten cKO than in the controls (p=0.0008) (Figure 6—figure supplement 1C,D). Enhanced OL accumulation in the cKO mice was also evident based on the number of BrdU+CC1+ cells (Figure 6—figure supplement 1D), indicating that the new OLs had been generated from proliferating OPCs.
Together, these results suggest that Pten-inactivated OPCs have a greater potential to regenerate new OLs than normal OPCs in the demyelinated CNS.
Although PI3K-Akt potentially regulates numerous downstream signaling cascades, mTOR is often thought to be the main downstream effector in OL development and myelination (Figlia et al., 2018; Wahl et al., 2014; Wood et al., 2013). However, our OPC-specific Tsc1 ablation (inducing mTORC1 hyperactivity) impaired OL development, whereas OPC-specific Pten-ablation (which also activates mTORC1) promoted new OL generation. Thus, these opposite results raise the possibility that the Pten-deletion-induced OL promotion is not mediated only by mTORC1 activity. While mTORC1 signaling is important for myelin growth, it is also possible that different PTEN-related signaling mechanisms regulate the OPC-to-OL lineage progression. Indeed, we observed that almost all pS6 (an mTORC1 activity indicator) immunoreactivities were localized at either EYFP+NG2- pre-OLs (about 25%) or EYFP+CC1+ OLs (more than 70%) in the CC of Pdgfra-CreER; R26-EYFP; ±Ptenf/f (P20 +21) mice. The percentage of NG2+ OPCs among pS6+ cells was negligible (less than 1%) even after Pten cKO (Figure 7A). These observations suggest that the actual site of mTORC1 activity of OL lineage is not NG2+ OPC, but more mature OL lineage cells.
In order to address the question of whether mTOR-dependent signaling mediates the enhanced OL generation from PTEN-ablated OPCs, we deleted both Mtor and Pten from OPCs with Pdgfra-CreER; R26-EYFP; ±Ptenf/f; ±Mtorf/f (Risson et al., 2009) mice at P20. First, the effectiveness of Mtor deletion was confirmed by a marked decrease of EYFP+pS6+ cells in Mtor or Pten-Mtor double cKO mice (Figure 7B,C). To our surprise, the degree of enhancement of OL generation was comparable between Pten cKO and Pten-Mtor double cKO mice (P20 +21) (Figure 7—figure supplement 1), indicating that the Pten cKO phenotypes were not reversed, even after mTOR was effectively inactivated.
On the other hand, after OPC-specific Mtor deletion in 4HT-administered Pdgfra-CreER; R26-EYFP; Mtorf/f mice (P20 +21), there was a marked reduction in EYFP+ cells compared to controls, particularly for CC1+ OLs, but not NG2+ OPCs (Figure 7D–G), confirming a critical role of mTOR in OL development (Wahl et al., 2014). However, most strikingly, new OL (EYFP+CC1+) generation was still greatly enhanced in the OPC-targeted Pten-Mtor double cKO mice, as compared to Mtor cKO mice (Figure 7D–F). Consequently, even without mTOR, the Pten deletion sufficiently increased the percentage of mature OLs among EYFP-labeled cells in the CC and CTX (Figure 7H). The Pten-deletion-induced OL promotion was also observed in the ventral GM area of the SC of Pten-Mtor cKO mice: OPC-specific Mtor cKO reduced new OL accumulation at P20 +21, but Pten-Mtor double cKO reversed this phenotype, or showed even more new OLs than in controls (Figure 7—figure supplement 2).
EM analyses confirmed that OPC-specific Pten-Mtor double cKO lowered the g-ratio to a level comparable to Pten cKO (Figure 7I–K), and formed new myelination for more axons (Figure 7L). These striking results suggest that one or more alternative signaling mechanisms to mTOR mediate the enhanced OL differentiation after OPC-specific PTEN inactivation.
GSK3β is another downstream target of activated Akt (Figure 8A), and it regulates cell differentiation, neural progenitor proliferation, and axonal growth in the CNS (Hur and Zhou, 2010). Interestingly, when cultured OPCs were treated by triiodothyronine (T3, 30 ng/ml) to induce OL differentiation, the levels of phospho-GSK3βSer9 were significantly increased by ~65% in 1 day (p=0.028, paired t-test), at which point the cultured cells are a mixture of pre-OLs (GalC+,O4+, MBP-) (Dai et al., 2014) and a small fraction of OPCs (weak PDGFRα+) (Figure 8B–D).
Although Akt-mediated phosphorylation at Ser9 inhibits GSK3β activity (Cross et al., 1995), it has not been shown that PTEN ablation leads to increased p-GSK3βSer9 levels in OL lineage cells (Goebbels et al., 2017). To determine whether PTEN ablation results in an increase of the inhibitory phosphorylation of GSK3β in OPCs, we isolated Pten-deleted EYFP+ cells from the Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice (P20 +21) by fluorescence-activated cell sorting (FACS). Western blot analysis of the isolated EYFP+ cells revealed that the levels of p-GSK3βSer9 were significantly increased by PTEN ablation (p=0.0138) (Figure 8E). It also appeared that there were increases in the levels of phosphorylated Akt (p-AktSer473) and p-GSK3βSer9 in the CTX of Olig1-Cre; Ptenf/f late (E18.5) embryos, compared with the Olig1-Cre embryos (Figure 8F). These observations provide compelling evidence that PTEN inactivation in OPCs leads to GSK3β inactivation in vivo. Moreover, in our OPC culture from the 4HT-administered Ptenf/f; ±Pdgfra-CreER (P1 +1) pups, there was a consistent tendency (but not significant, n = 3 replicates) of increase of p-AktSer473 (Figure 8G) and p-GSK3βSer9 (Figure 8H) in Pten cKO OPCs.
Next, we examined the effects of OPC-specific GSK3β inactivation in vivo as a likely signaling event after PTEN ablation. We bred Pdgfra-CreER; R26-EYFP mice with Gsk3bf/f mice (Patel et al., 2008), and injected tamoxifen into Pdgfra-CreER; R26-EYFP; ±Gsk3bf/f mice between P20 and P22 (Figure 9A). We isolated EYFP+ cells from these mice at P20 +8, and performed RT-qPCR for Gsk3b mRNA levels. The results of RT-qPCR showed that there was a reduction of Gsk3b mRNA by more than 90% in EYFP+ cells, confirming an effective Gsk3b ablation in OPCs (Figure 9B).
The tamoxifen-administered Pdgfra-CreER; R26-EYFP; ±Gsk3bf/f were also examined 21 days later (P20 +21). The last 4-day-BrdU chasing (see Figure 9A) revealed significantly enhanced OPC proliferation in the Gsk3b cKO mice, which is similar to the OPC-specific Pten cKO mice (Figure 9C). Moreover, there were marked increases in both total EYFP+ cells and EYFP+CC1+ OLs in both the CC and CTX of Gsk3b cKO mice (Figure 9D–F), suggesting enhanced OL generation. In contrast, densities of EYFP+NG2+ OPCs and total OPCs were not altered (Figure 9G,H), and as a consequence, the percentage of OLs or that of OPCs among EYFP+ OL lineage cells was altered (Figure 9I). EM analyses also confirmed that myelin thickness and the degree of new axon myelination were significantly enhanced by Gsk3b cKO (P13 +17) (Figure 9J–M). These results strongly suggest that resting GSK3β activity inhibits OPC-to-OL lineage progression in the brain, and similar to PTEN ablation, OPC-specific GSK3β inactivation promotes OL development.
We also targeted Gsk3b specifically in OLs with Mog-iCre; ±Gsk3bf/f mice (P31). Similar to OL-specific Pten cKO mice, there was no change in numbers of CC1+ OL and OPCs (Figure 9—figure supplement 1A,B), but the levels of MBP and CNP were increased in the OL-specific Gsk3b cKO mice (Figure 9—figure supplement 1C,D). These results suggest that GSK3β, like PTEN, negatively regulates not only OPC-to-OL transition but also myelin protein synthesis.
PI3K-Akt-mTORC1 signaling is critical for early myelination (Figlia et al., 2018; Wood et al., 2013), but precise control of mTORC1 activity is highly important for the progression of cell maturation and myelination. The results from OPC-specificTsc1 cKO mice in this study and other previous results (Jiang et al., 2016; Lebrun-Julien et al., 2014) suggest that mTORC1 hyperactivity either causes OL lineage arrest or is detrimental to OL survival. Similarly, Figlia et al. (2017) showed that Tsc1 cKO-induced mTORC1 hyperactivity delays the onset of Schwann cell myelination in the PNS, although residual mTORC1 activity derives myelin growth at later stages. Therefore, a large body of evidence suggests that even though mTORC1 is necessary, uncontrolled stimulation of mTORC1 activity may not be a practical approach for adult myelin repair.
This study provides the first in vivo genetic evidence that, in parallel with PI3K-Akt-mTORC1 signaling, PTEN-Akt-GSK3β signaling negatively regulates OL development and myelination. In particular, we demonstrate that PTEN-dependent inhibition of OPC-to-OL progression widely occurs both in the brain and the SC, and is persistent into adulthood. Our findings reveal that inactivation of PTEN in OPCs mobilizes its cell-intrinsic downstream signaling via mTOR-independent mechanisms, probably by controlling GSK3β activity. Thus, OL regeneration may be effectively enhanced by OPC-targeted inactivation of PTEN or GSK3β in the adult CNS.
Previously, PTEN has been only thought of as a negative regulator of OL membrane outgrowth (Goebbels et al., 2010; Harrington et al., 2010; Maire et al., 2014). However, neither its OPC-specific role nor its inhibitory action on the OPC-to-OL transition had been revealed. Moreover, it was thought that PTEN/Akt-initiated myelin changes are mediated mostly by mTORC1 (Goebbels et al., 2010; Narayanan et al., 2009). It is likely that the major differences between our results and those of earlier studies largely stem from the cell specificity of employed Cre drivers. Use of Cnp or Plp1 promoters for Cre or CreER expression (Goebbels et al., 2010) may have targeted mature OL-resident Pten. Similarly, a constitutively active form of mutant Akt was overexpressed in OLs with Plp1 promoter (Flores et al., 2008). Thus, those studies may not have been able to observe Akt-activated OPCs and the subsequent new OL development.
We used Pdgfra-CreER and Cre reporter mice to test the effects of OPC-targeted Pten deletion, and selectively followed the fates of affected OPCs. Although OPC-targeted gene deletion inevitably impacts OPC-derived new OLs and subsequent myelination at later time points, the phenotypic differences shown here clearly point to OPC-related mechanisms. Indeed, OL-specific Pten deletion with Mog-iCre in this study did not recapitulate OPC-specific cKO phenotypes (e.g. increases in OL number and percent of myelinated axons). Conversely, it is highly likely that the thickened myelin that was commonly observed in previous studies (Flores et al., 2008; Goebbels et al., 2010), as well as in the current study, is an OL-specific phenotype.
Pten has been also deleted in all Olig2-expressing cells by using Olig2-Cre mice, with either no change (Harrington et al., 2010) or a decrease (Maire et al., 2014) in Olig2+ cells, but with thickened myelin (Harrington et al., 2010; Maire et al., 2014). However, because Olig2 is expressed in subsets of neural stem/progenitor cell populations through embryonic development, it appears that not only OL lineage cells, but also astrocytes and some types of neurons, are affected by Olig2-Cre (Maire et al., 2014; Masahira et al., 2006). Thus, it is possible that the phenotypes of Olig2-Cre; Ptenf/f mice were complicated by non-cell-autonomous mechanisms (Codeluppi et al., 2009; Goebbels et al., 2017). More importantly, prolonged PTEN deficiency in OLs may have led to continuous myelin growth, which may eventually have exerted detrimental effects on the integrity of OLs and axons (Goebbels et al., 2010; Harrington et al., 2010; Narayanan et al., 2009), masking the positive impact of short-term PTEN inhibition on OPCs that was revealed in the present study.
PTEN is a cell-intrinsic regulator critical for axonal regeneration, and its deletion in neurons enhances post-injury axon regeneration (Liu et al., 2010; Park et al., 2008). A recent study has shown that PTEN-inactivated neurons are even able to recruit OPCs, and induce ectopic myelination in the cerebellar granule cell layer, in which myelination is normally absent (Goebbels et al., 2017). Our findings on PTEN’s role in OPCs, along with those previous observations, have broaden our understanding of PTEN function in CNS neural repair and regeneration. One concern in targeting PTEN for better myelin repair is that long-term PTEN inhibition (or Akt over-activation) in OLs may cause axonal pathologies (Flores et al., 2008; Harrington et al., 2010). However, if PTEN inactivation were effectively guided to OPCs via short-term pharmacological delivery near the site of CNS injury, this could recruit and activate OPCs toward OL differentiation, while minimizing systemic complications. A recent ex vivo approach demonstrated a similar rationale by inhibiting PTEN in cultured OPCs (De Paula et al., 2014).
It is important to identify the downstream mechanism that mediates PTEN-inactivation-enhanced OL generation. PTEN inhibition results in phosphatidylinositol (3,4,5)-trisphosphate (PIP3) accumulation, and thus activates Akt. Even though Akt is known to phosphorylate many different downstream molecules, its major targets may differ according to the type of cell. TSC1/2 phosphorylation and subsequent mTORC1 activity has been considered the most responsible mediator of PTEN-inactivation-dependent axon regeneration (Liu et al., 2010; Park et al., 2008) and PTEN-inactivation-induced myelin membrane outgrowth (Goebbels et al., 2010; Harrington et al., 2010; Narayanan et al., 2009). In the PNS, PTEN also regulates the onset of myelination, exclusively through mTORC1 activity (Figlia et al., 2017). In most cases, the supporting evidence for the importance of mTORC1 has been obtained with rapamycin, an mTORC1 inhibitor (Du et al., 2015; Figlia et al., 2017; Goebbels et al., 2010; Narayanan et al., 2009; Park et al., 2008). However, given the widespread impact of systemically administered rapamycin on multiple cell types (Goldshmit et al., 2015; Srivastava et al., 2016), the in vivo results may be difficult to interpret in terms of cell-type-specific mechanisms. We selectively deleted Mtor in OPCs, and subsequent cell fate analysis confirmed the importance of mTOR in OL development, in agreement with prior studies (Bercury et al., 2014; Lebrun-Julien et al., 2014; Wahl et al., 2014). It is noteworthy that the degree of impairment of OL development shown by OPC-specific Mtor ablation in this study was far greater than that observed with rapamycin administration (Narayanan et al., 2009). However, despite this effective mTOR inactivation, it did not reverse PTEN-inactivation-promoted oligodendrogenesis, indicating that the observed OL promotion was not mediated by mTOR signaling. Furthermore, when PTEN and mTOR were deleted together, the activated signaling mechanisms in OPCs sufficiently compensated for the loss of mTORC1, stimulating OL generation more than in normal OL development. Therefore, our results reveal that there are one or more PTEN-downstream mechanisms besides Akt-mTORC1 signaling that are critical for OL development and CNS myelination.
In the present study, we have shown that Pten ablation in OPCs also leads to the inhibitory GSK3β phosphorylation (at Ser9). Therefore, it is likely that PTEN inactivation results in GSK3β inhibition in OL lineage cells. Finally, we show that OPC-specific Gsk3b deletion also promotes OL development in the brain similar to that of Pten ablation. Interestingly, it has been shown that PTEN-inhibition (or PI3K activation) induces sensory axon regeneration, not via mTORC1 signaling (Christie et al., 2010; Saijilafu et al., 2013), but via GSK3α/β activity inhibition in vitro (Saijilafu et al., 2013). Moreover, Akt-dependent CNS axon regeneration is further promoted by GSK3β inhibition, in addition to mTORC1 signaling-dependent mechanisms (Miao et al., 2016), suggesting an important role of Akt-GSK3β signaling in axonal regeneration. A recent study has shown that a local delivery of GSK3β inhibitors promoted OL development from stem cells in the subventricular zone (Azim and Butt, 2011), although opposite in vitro results have also been also reported (Zhou et al., 2014). Therefore, an effective OPC-targeted PTEN or GSK3β inhibition may have therapeutic potential and benefit for expedited remyelination.
In summary, our study reveals that OPC-resident PTEN persistently regulates OL lineage progression, and demonstrates a beneficial potential of PTEN inhibition for better OL regeneration. Moreover, it suggests that mTORC1-dependent signaling is not the only major downstream player for these positive responses, but that PTEN-Akt-GSK3β signaling poses a negative balance, and plays an important role in OL lineage progression. These results support the idea that OPC-targeted cell-specific genetic approaches are powerful tools for identifying and evaluating cell-intrinsic molecules for the promotion of OL regeneration.
Mice homozygous for floxed alleles of Tsc1 (Uhlmann et al., 2002) (STOCK Tsc1 <tm1Djk>/J, RRID:IMSR_JAX:005680), Pten (Lesche et al., 2002) (STOCK Pten <tm1Hwu>/J, RRID:IMSR_JAX:006440), Mtor (Risson et al., 2009) (STOCK Mtor <tm1.2Koz>/J, RRID:IMSR_JAX:011009), ROSA26-EYFP (Srinivas et al., 2001) (STOCK Gt(ROSA)26Sor < tm1(EYFP)Cos>/J, RRID:IMSR_JAX:006148) and ROSA26-mEGFP (Muzumdar et al., 2007) (STOCK Gt(ROSA)26Sor < tm4(ACTB-tdTomato,-EGFP)Luo>/J, RRID:IMSR_JAX:007576) were purchased from The Jackson Laboratory. Olig1-Cre (Lu et al., 2002) mice were obtained from Dr. David Rowitch (UCSF), and are now available from The Jackson Laboratory (STOCK 129S4-Olig1 < tm1(cre)Rth>/J, RRID:IMSR_JAX:011105). Gsk3bf/f mice (Patel et al., 2008) were obtained from Dr. Yang Hu (Stanford University), and are now available from The Jackson Laboratory (STOCK Gsk3b < tm2Jrw>/J, RRID:IMSR_JAX:029592). Pdgfra-CreER mice (Kang et al., 2010) (STOCK Cg-Tg(Pdgfra-cre/ERT)467Dbr/J, RRID:IMSR_JAX:018280) and Mog-iCre mice (Buch et al., 2005) have been described before. The Mobp-EGFP mice were developed by GENSAT, and are available from MMRRC (STOCK Tg(Mobp-EGFP)IN1Gsat/Mmucd, RRID:MMRRC_030483-UCD).
Genotyping was performed using the PCR conditions recommended by The Jackson Laboratories. The genomic background of all multiple transgenic mice used here should be considered as B6SJL, C57BL/6 and 129 mixed. For all OPC fate analyses with conditional target gene deletion, age-matched relative Pdgfra-CreER; R26-EYFP (or Pdgfra-CreER; R26-mEGFP) mice were used as the control. All western blot analysis for myelin proteins, littermate homozygous mice for floxed gene (without Cre) were used as the control. For regular EM analyses, age-matched Ptenf/f or Ptenf/f; Mtorf/f mice (without Cre, but some contained R26-EYFP assuming that it is not expressed without Cre) were used as the control. The number of the mice and genotype information are shown in individual source data files that are linked to figure legends. All animal procedures were conducted in compliance with animal protocols approved by the Institutional Animal Care and Committee (IACUC) of Temple University School of Medicine.
(Z)−4-Hydroxytamoxifen (4HT, sigma H7904) was prepared as described before (Badea et al., 2003; Kang et al., 2010). One mg of 4HT was administered per intraperitoneal (i.p.) injection, and up to two injections were given per day with at least 6 hr apart. For several sets of experiments, tamoxifen (sigma T5648) was prepared according to Madisen et al. (2010) and injected in a similar way to that of 4HT. Individual tamoxifen injection dose was 40 mg/kg. The total number of 4HT (or tamoxifen) injections and the first injected age for each experiment are stated in results and/or figure legends.
For cell proliferation analysis, mice received 5-bromo-2'-deoxyuridine (BrdU) injections (50 mg/kg per injection, i.p.) twice a day, in addition to BrdU-containing drinking water (1 mg/ml BrdU in tab water with 1% sucrose supplement) for the last 4 days before mouse sampling (i.e., seven injections from P38 to P41). For the experiment of long-term OPC fate analysis with control and Pten cKO mice (P20 +70) (Figure 3), the BrdU was injected between P80 and P84, (a total of 10 injections without BrdU drinking water). For lysolecithin-induced SC demyelination (Figure 6—figure supplement 1), BrdU was administered only via drinking water for 13 days after LPC injection.
Eight days after the first tamoxifen injection was given to Pdgfra-CreER; R26-EYFP; ±Ptenf/f mice (i.e. P58), the mice were anesthetized with Ketamine (120 mg/kg) and Xylazine (8 mg/kg). After mice were fixed on a stereotaxic instrument, an L-α-Lysolecithin (L-α-Lysophosphatidylcholine or LPC in saline; Sigma L4129)-loaded Hamilton syringe was inserted into right CC (AP, +0.8 mm; ML, −0.8 mm; DV, −2.0 mm from bregma) (Franklin and Paxinos, 2008). The Hamilton syringe was equipped with a 33-gauge needle (45° beveled tip) attached to a motorized stereotaxic injector (Stoelting). One μl of 1% LPC was infused at a rate of 0.05 μl/min. After injection, the needle was held in place for an additional 5 min before removal.
For spinal cord demyelination, the spinal cord was exposed between lumbar spinal segment 4 and 5 (L4/L5), and stabilized with clamps (Di Maio et al., 2011). A 0.5 μl of 1% LPC was injected 0.5 mm deep into the dorsal column WM of the spinal cord on the right side of midline dorsal artery at a rate of 0.1 μl/min with an angle of 45 degree by using a glass micropipette and a micromanipulator and digital injector (World Precision). The needle was left in the place for additional two min to avoid backflow of injected LPC. A total of three injections were made into sites spanning the L4/L5. After recovery from anesthesia, buprenorphine (0.7 mg/kg) was administrated by a subcutaneous injection.
Mice were deeply anesthetized with sodium pentobarbital (100 mg/kg, i.p.) and briefly perfused with PBS followed by 4% paraformaldehyde (PFA in 0.1 M phosphate buffer, pH 7.4). Brains and the spinal cords were isolated and were subjected to post-fixation in 4% PFA for additional 6 hr at 4°C. The sampled tissues were incubated in 30% sucrose solution (in PBS) for at least 36 hr at 4°C for cryoprotection, and were embedded in Optimal Cutting Temperature (OCT) medium and frozen on dry ice. Thirty-five µm-thick sections were prepared using cryostat, and collected in PBS.
The information of antibodies and key chemical reagents used in this study is also shown in the key resources table. Tissue sections were rinsed in PBS for 5 min (twice), permeabilized in 0.3% Triton X-100 (in PBS) for 10 min, and blocked (0.3% Triton X-100, 5% normal donkey serum, 1% BSA in PBS) for 1 hr at room temperature (RT). For fixed OPC culture, coverslips were permeabilized in 0.1% Triton X-100 (in PBS) for 20 min and blocked (0.1% Triton X-100, 5% normal donkey serum). The sections were then incubated with the primary antibodies in the same blocking solution in a free-floating manner overnight at 4°C. The primary antibodies (Abs) used in the present study were rabbit anti-GFP (Frontier Institute, #GFP-Rb-Af2020, RRID:AB_2491093, 1:500 or 1:1,000 for TSA amplification), rabbit anti-GFP (Proteintech, #50430–2-AP, RRID:AB_11042881, 1:500), goat anti-GFP (Frontier Institute, #GFP-Go-Af1480, RRID:AB_2571574, 1:500), mouse anti-APC (clone CC1, EMD Millipore, #OP80, RRID:AB_2057371, 1:100), guinea pig anti-NG2 (a gift from Dr. Dwight Bergles at Johns Hopkins University, 1:4,000), mouse anti-MBP (clone SMI-99, Covance, #SMI-99P, RRID:AB_10120130, 1:500), rabbit anti-phospho-S6Ser240/Ser244 (Cell Signaling Technology, #2215, RRID:AB_331683, 1:500), rat anti-BrdU (Accurate, #OBT0030G, RRID:AB_609567, 1:500), and rabbit anti-Olig2 (EMD Millipore, #AB9610, RRID:AB_570666, 1:500), rabbit anti-Ki67 (Abcam, #ab15580, RRID:AB_443209, 1:500), anti-mouse GalC (Millipore, #MAB342, RRID:AB_2073708, 1:100), rat anti-PDGFRα (BD Biosciences, #558774, RRID:AB_397117, 1:250). After washing with PBS for 5 min (three times), the sections were incubated with secondary antibodies for 2 hr at RT. The secondary Abs Alexa Fluor 488-conjugated donkey anti-rabbit (1:500) or anti-goat (1:500), Cy3-conjugated donkey anti-mouse, anti-rabbit or anti-rat (1:500), Alexa-Fluor 647-conjugated donkey anti-guinea pig (1:500), and HRP-conjugated donkey anti-rabbit (1:1,000) Abs (Jackson ImmunoResearch). In all cases nuclei were co-stained with DAPI. Sections were washed in PBS for 5 min (three times) and mounted on slides with ProLong Diamond antifade mounting media (Thermo-Fisher).
For BrdU staining, prior to blocking, brain or spinal cord sections were pre-treated with 2 N HCl for 30 min at 37°C, followed by neutralization with 0.1 M sodium borate buffer (pH 8.5) for 10 min (twice). For BrdU and EYFP co-immunostaining, brain or spinal cord sections were pre-treated with 0.3% H2O2 for 30 min at RT before blocking. Anti-EGFP immunostaining and BrdU (plus other immunostaining) staining were carried out sequentially. Anti-EGFP staining was performed first and the signal was further amplified with TSA plus immunofluorescence kit (PerkinElmer) according to manufacturer’s instruction.
As the gray and white matter in the brain and spinal cord have different rates of new OL generation depending on developmental stage or examined fate mapping time window (Kang et al., 2010), the degree of new OL generation was analyzed for both gray and white matter areas in the CNS. Fluorescence images were captured from the motor cortex and corpus callosum from the brain, and gray and white matter from the ventral spinal cord with a Zeiss epifluorescence microscope (Axio-Imager M2) and Axiovision 7.0 or Zen software (Zeiss, Germany). Three or four sections were used per mouse, and at least three mice were used per group. Confocal images were captured with 1 μm step with a laser scanning confocal microscope (Leica TCS SP8) and LAS X software.
Cortices were isolated from mice and snap frozen in dry ice and were stored at −80°C until use. Lysates were prepared by homogenizing the cortices in RIPA buffer (50 mM Tris pH 7.5, 150 mM NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 1% NP-40) supplemented with protease inhibitor cocktail (Thermo-Fisher). Freshly sorted cells or culture were also lysed with RIPA buffer (1000 cells per µl). Protein concentration was determined using BCA protein assay kit (Thermo-Fisher). Twenty µg of protein or 10 µl of cell lysate was mixed with 2X Laemmli Sample Buffer (Bio-Rad), and were resolved with 4–15% gradient Bis-Tris gel (Bio-Rad) or 12% Bis-Tris gel by electrophoresis. Proteins were transferred to PVDF membranes (EMD Millipore). Membranes were blocked in 5% skim milk for 1 hr at RT, and were incubated with Abs for 1 hr at RT. We used mouse anti-PLP (EMD Millipore, #MAB388, RRID:AB_177623, 1:1000), rabbit anti-MBP (Cell Signaling Technology, Cat # 78896, 1:2000), rabbit anti-CNPase (Phosphosolutions, #325-CNP, RRID:AB_2492062, 1:2000), mouse anti-MAG (Santa Cruz Biotechnology, #sc-166849, RRID:AB_2250078, 1:2000), mouse anti-MOG (Santa Cruz Biotechnology, #sc-166172, RRID:AB_2145540, 1:2000), rabbit anti-GSK3β (Cell Signaling Technology, Cat # 12456, 1:2000), rabbit anti-phospho-GSK3βSer9 (Cell Signaling Technology, #9323, RRID:AB_2115201, 1:2,000), rabbit anti-Akt (Cell Signaling Technology, #4691, RRID:AB_915783, 1:2000), rabbit anti-phospho-AktSer473 (Cell Signaling Technology, #4060, RRID:AB_2315049, 1:2000), and mouse anti-β-actin (Santa Cruz Biotechnology, #sc-47778, RRID:AB_2714189, 1:2000) Abs. After brief washing with TBS-T for 10 min (three times), the membranes were incubated with HRP-conjugated anti-rabbit or anti-mouse IgG (Jackson ImmunoResearch, 1:20,000) in 5% skim milk for 1 hr at RT. Chemiluminescent signals were detected with ECL kit (Supersignal West Dura, Thermo-Fisher). Protein bands were quantified with Image J software. In case of sequential probing with different Abs, membranes were stripped with stripping solution (2% SDS, 0.06 M Tris-HCl, pH 6.8, 0.8% 2-mercaptoethanol) for 10 min at 55°C, and washed with TBS-T three times before a subsequent probing. The full-length original western blot images are shown in source data files linked to the relevant figure legends.
For a western for PLP, protein samples were resolved with 4–15% gradient Bis-Tris gel (Bio-Rad). The PVDF membrane was blocked in Odyssey Blocking Buffer (PBS) (LI-COR) for 30 min at RT, and probed with anti-PLP and anti-β-Actin Abs overnight at 4°C. After brief washing with TBS-T, membranes were incubated with IR-Dye 800 CW goat anti-mouse (LI-COR, 1:5,000) in the same blocking buffer that was supplemented with 0.02% Tween, 0.01% SDS for 1 hr at RT. Protein bands were detected with Odyssey infrared scanner (LI-COR). Band intensity was quantified with Image studio (version 3.1) software.
To isolate PTEN-deficient cells from the Pten cKO mice, we used tamoxifen-administered Pdgfra-CreER; R26-EYFP; ±Ptenf/f (P20 +21) mice. Cerebral cortices were dissected and processed according to manufacturer’s instructions (Neuronal Tissue Dissociation Kit (P), 130-092-628, Miltenyi Biotec). To remove myelin debris, the dissociated cells were resuspended in ice-cold 0.9 M sucrose solution (in HBSS, without calcium and magnesium) and centrifuged at 300 × g for 10 min. After myelin suspension was discarded, the pellets were resuspended in 1 ml of fresh HBSS complemented with 1 mM EDTA and 1% BSA. Cells were sorted using (BD FACSAria IIµ) at the Flow Cytometry Core of Temple University. The sorted EYFP+ cells from control and Pten cKO mice were used for western blot analysis. For the validation of Gsk3b deletion, EYFP+ cells were also isolated from whole forebrains of the tamoxifen-administered Pdgfra-CreER; R26-EYFP; ±Gsk3bf/f mice (P20 +8) by FACS. The Gsk3b-deleted cells were subjected to RNA extraction.
Total RNA was extracted from EYFP+ cells with RNeasy Plus Micro Kit (Qiagen). The first strand cDNA was synthesized using SuperScript III First-Strand Synthesis kit (Invitrogen). The RT-qPCR was performed with QuantiTect SYBR green (Qiagen) on the Applied Biosystems Step One Plus RT-PCR system. The thermal cycle profile was: 95°C for 15 min, 40 cycles of 95°C for 15 s, 55°C for 30 s and 72°C for 15 s. The primers used for Gsk3b mRNA detection were: Gsk3b Ex1 (F)- 5’ GACCGAGAACCACCTCCTTT 3’ and Gsk3b Ex2 (R)- 5’ ACTGACTTCCTGTGGCCTGT 3’. The primers for β-Actin (Actb) were: b-actin (F)−5’ TGACAGGATGCAGAAGGAGA 3’ and b-actin (R)- 5’ cgctcaggaggagcaatg 3’.
Mice were deeply anesthetized with pentobarbital (100 m/kg, i.p) and perfused with PBS followed by 2.5% PFA, 2% glutaraldehyde (in 0.1 M phosphate buffer, pH 7.4) for regular electron microscopy (EM). After tissue isolation, samples were post-fixed for 4 hr at 4°C, and transferred to 0.1 M phosphate buffer at 4°C. The tissues were dehydrated using a graded ethanol series and embedded in EMbed-812 (EMS). Thin sections were stained with uranyl acetate and lead citrate. The obtained sections were visualized with an electron microscope (JEOL 1010 electron microscope fitted with a Hamamatsu digital camera) at 12,000X magnification. The image acquisition was performed with AMT Advantage image capture software and g ratio measurement with image J.
For immuno-EM, 4% PFA/0.1% glutaraldehyde (in 0.1 M phosphate buffer) was used for perfusion. Brains were removed from the skull and post-fixed in the same fixative for 4 hr. Microslicer sections containing the corpus callosum (60 µm in thickness) were incubated with anti-EGFP Ab (Frontier Institute) for 12 hr followed by the incubation with Nanogold-conjugated anti-rabbit IgG (Nanoprobes, Cat # 2003, 1:100) for 2 hr at RT. Immunogold was intensified with the HQ Silver Enhancement kit (Nanoprobes). The sections were post-fixed with 2% osmium tetroxide, stained with 2% uranyl acetate, dehydrated, and embedded in epoxy resin. Ultrathin sections (70 nm in thickness, Leica Ultracut) were examined using an electron microscope (H-7650, Hitachi).
OPC primary culture was prepared from WT mice, or Pten cKO and their littermate control mice according to Chen et al. (2007). For Pten cKO mice, each P1 pup of a whole litter of Ptenf/f;±Pdgfra-CreER mice received a single 4HT injection (0.2 mg s.c.) at P1. At P2, genotyping was performed for the litter, and the pups were killed 24 hr after 4HT injection (i.e., P1 +1). Cortices were isolated and digested with papain and DNase I, followed by mechanical dissociation. Cells were re-suspended in DMEM supplemented with 10% horse serum and 1% penicillin/streptomycin. The mixed glial cells were seeded onto T75 flask coated with poly-D-lysine, and were maintained for 7–10 days. OPCs were further enriched by differential shaking and were seeded onto poly-D-lysine-coated coverslips. Cells were kept in defined medium. To induce OPC differentiation, the culture media was supplemented with triiodothyronine (T3, 30 ng/ml). For western blot, protein lysate was prepared in RIPA buffer 24 hr later. For immunocytochemistry, the cells were fixed with 4% PFA for 15 min and rinsed three times in PBS.
Data analysis was performed with GraphPad Prism 5.0 software. Unless specified otherwise in the figure legends, most statistical analyses were carried out using two-tailed unpaired Student’s t-test for two group comparison, or One-way ANOVA with Bonferroni post-hoc correction for statistical evaluation of more than two groups. For the western blot analyses for GSK3βSer9 or AktSer473 with OPC or pre-OL culture, paired student t-test was used to compare OPC and pre-OL, or control and Pten cKO culture lysates.
The data and error bars indicate mean ±standard error of the mean (S.E.M.), where ‘n’ refers to the number of mice or replicates of culture. Prior power analyses were not performed to calculate the sample size for the experiments, because the size of the effects expected was unknown before doing the experiments. However, sample sizes were estimated based on similar experiments reported in the relevant literature in the field (Figlia et al., 2017; Jiang et al., 2016). The p values, S.E.M, t values, and d.f. of each set of comparisons are shown in source data files.
Hypomyelination following deletion of Tsc2 in oligodendrocyte precursorsAnnals of Clinical and Translational Neurology 2:1041–1054.https://doi.org/10.1002/acn3.254
Isolation and culture of rat and mouse oligodendrocyte precursor cellsNature Protocols 2:1044–1051.https://doi.org/10.1038/nprot.2007.149
PTEN inhibition to facilitate intrinsic regenerative outgrowth of adult peripheral axonsJournal of Neuroscience 30:9306–9315.https://doi.org/10.1523/JNEUROSCI.6271-09.2010
The Rheb-mTOR pathway is upregulated in reactive astrocytes of the injured spinal cordJournal of Neuroscience 29:1093–1104.https://doi.org/10.1523/JNEUROSCI.4103-08.2009
Pten Deletion Promotes Regrowth of Corticospinal Tract Axons 1 Year after Spinal Cord InjuryJournal of Neuroscience 35:9754–9763.https://doi.org/10.1523/JNEUROSCI.3637-14.2015
Myelin regeneration: a recapitulation of development?Annual Review of Neuroscience 34:21–43.https://doi.org/10.1146/annurev-neuro-061010-113629
Constitutively active Akt induces enhanced myelination in the CNSJournal of Neuroscience 28:7174–7183.https://doi.org/10.1523/JNEUROSCI.0150-08.2008
The Mouse Brain in Stereotaxic Coordinates (3rd edition)New York: Academic press.
Intracellular signaling pathway regulation of myelination and remyelination in the CNSExperimental Neurology 283:501–511.https://doi.org/10.1016/j.expneurol.2016.03.008
Rapamycin increases neuronal survival, reduces inflammation and astrocyte proliferation after spinal cord injuryMolecular and Cellular Neuroscience 68:82–91.https://doi.org/10.1016/j.mcn.2015.04.006
Amino acid sufficiency and mTOR regulate p70 S6 kinase and eIF-4E BP1 through a common effector mechanismJournal of Biological Chemistry 273:14484–14494.https://doi.org/10.1074/jbc.273.23.14484
Oligodendrocyte PTEN is required for myelin and axonal integrity, not remyelinationAnnals of Neurology 68:703–716.https://doi.org/10.1002/ana.22090
Balanced mTORC1 activity in oligodendrocytes is required for accurate CNS myelinationJournal of Neuroscience 34:8432–8448.https://doi.org/10.1523/JNEUROSCI.1105-14.2014
PTEN deletion enhances the regenerative ability of adult corticospinal neuronsNature Neuroscience 13:1075–1081.https://doi.org/10.1038/nn.2603
Akt signals through the mammalian target of rapamycin pathway to regulate CNS myelinationJournal of Neuroscience 29:6860–6870.https://doi.org/10.1523/JNEUROSCI.0232-09.2009
Myelination of the nervous system: mechanisms and functionsAnnual Review of Cell and Developmental Biology 30:503–533.https://doi.org/10.1146/annurev-cellbio-100913-013101
Polydendrocytes (NG2 cells): multifunctional cells with lineage plasticityNature Reviews Neuroscience 10:9–22.https://doi.org/10.1038/nrn2495
Tissue-specific role of glycogen synthase kinase 3beta in glucose homeostasis and insulin actionMolecular and Cellular Biology 28:6314–6328.https://doi.org/10.1128/MCB.00763-08
Muscle inactivation of mTOR causes metabolic and dystrophin defects leading to severe myopathyThe Journal of Cell Biology 187:859–874.https://doi.org/10.1083/jcb.200903131
Beth StevensReviewing Editor; Boston Children's Hospital, Harvard Medical School, United States
In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.
Thank you for submitting your article "PTEN regulates the cell lineage progression from NG2+ glial progenitor to oligodendrocyte via mTOR-independent signaling" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom, Beth Stevens is a member of our Board of Reviewing Editors and the evaluation has been overseen by a Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Teresa Wood (Reviewer #2).
The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.
Gonzalez-Fernandez and colleagues examine the role of PTEN signaling in oligodendrocyte progenitor proliferation and differentiation into mature oligodendrocytes during development and early adulthood, both in the context of a healthy CNS and during remyelination. These data are important to the myelin field as recent research has focused on the role of PI3K/Akt/mTOR signaling in myelination. Their findings point to an mTOR-independent role for PTEN through GSK3β. Mice lacking PTEN specifically in OPCs exhibited continued differentiation of OPCs into mature OLs, even in early adulthood. This corresponds with increased myelin production, which is also observed in OL-specific deletion of PTEN. Interestingly, loss of mTOR did not abrogate these effects of PTEN deletion, suggesting an mTOR-independent mechanism. Additional data demonstrate instead that the effects of PTEN deletion require GSK3β.
The data are clearly presented and make important and novel contributions; however, there are several experiments which would strengthen the conclusions as summarized in essential revisions below. In addition to these revisions, there are several additional points raised by each reviewer that need to be clarified and addressed in a revised paper as summarized at the end.
1) The authors state that their Tsc1 data indicate decreased OL survival, yet do not show the data to support this statement. Is the decreased mature OL number is a result of decreased proliferation and/or differentiation? To distinguish these possibilities the authors could co-stain for NG2/caspase, CC1/caspase, and/or NG2/Ki67
2) In Figure 4, the authors examine PLP and MBP by Western blotting, and find that PLP is significantly increased but MBP is not. Given the different results, it would be more convincing of the structural integrity of increased myelin thickness if other myelin proteins were examined such as MOG, CNP, and MAG.
3) The authors show that increased mature OL numbers in PTEN cKO mice correspond to increased myelin thickness and number of myelinated axons. Is there any evidence to suggest that decreased OL numbers in TSC1 cKO mice correlate with decreased myelination, either by immuostaining or Western Blot techniques?
4) Is GSK3 responsible only for the differentiation enhancement that was observed in PTEN cKO mice, or is it also responsible for myelin production by mature OLs? Performing either immunostaining/Western blotting for myelin proteins in GSK3 cKO mice could address this. Is it possible that PTEN mediates differentiation via GSK3 but myelin production via a different mTOR independent mechanism?
5) Much of the data in the paper is based on the calculations of EYFP positive cells, which is taken as a reporter of Cre-recombinase activity. Another possibility is that knockout of the gene of interest in each could affect expression levels of EYFP protein. One can imagine a scenario where Cre-recombinase is active but EYFP is not detected. The authors could stain for Cre expression and see if they see it in cells with no EYFP.
6) Figure 1: Tsc1 deletion (which hyperactivates mTOR) in OPCs resulted in a decrease in the number of CC1+ cells, in Figure 6, the deletion of mTOR in OPCs had the same effect – less number of CC1+ cells. How do the authors interpret the result that Tsc1 deletion and mTOR deletion have the same effect? Moreover, how does the g-ratio and number of myelinated axons change with the deletion of mTOR alone?
7) The effect of GSK3β on OL development and myelination is shown nicely; however, the connection between PTEN-Akt and Gsk3β still needs to be clarified. The authors also mention this missing information in the Discussion section. In order to support this link more with the resources they have, they can check how the Gsk3β activity (phosphorylation by Akt) and the activity/level of the proteins downstream of Gsk3β change in Ptenf/f mouse.
8) Validation of the GSK3βf/f mouse doesn't appear in the paper and should be included.
9) For the g-ratio calculations the authors should perform statistics on the biological replicates and not the individual axons counted. Can the authors please confirm that this is the case?
Fernandez and colleagues investigate a well-known pathway that regulates myelin formation and thickness in the white matter development, mTOR. It has been known for some time that mTOR signaling regulates developmental myelination, but most experiments have examined this pathway in the context of development and not in the adult or injured brain. Here the authors performed experiments where they ablated upstream negative regulators of mTOR signaling, Tsc1 and PTEN, and examined their effects on OPC differentiation in the adult brain and following focal demyelination using lysolecithin.
The authors show that deletion of Tsc1 in OPCs leads to decreased numbers of oligodendrocyte-lineage cells (EYFP+) in the adult white matter as well as a reduction of the percentage of OLs and increase in the percentage of OPCs. Deletion of Pten lead to modest increases in oligodendrocyte lineage cell number as well as modest increases in OL generation at the expense of OPCs when induced with tamoxifen at P45. This was accompanied by increases in myelin production, but none of these effects were observed in an oligodendrocyte Pten conditional knockout. OPC ablation of Pten also leads to accelerated OL regeneration after LPC demyelination. Finally, the authors show that mTOR signaling is dispensable for these phenotypes in the Pten OPC cKO and that deletion of another downstream signaling pathway, GSK3, phenocopies the Pten OPC cKO.
These experiments show a critical divergence of Pten/mTOR signaling in OPCs versus OLs which could be important in demyelinating diseases where OPC differentiation is delayed or inefficient. This study offers a mechanistic look at the effects of different pathway components on OPC differentiation and myelination using several different conditional knockout lines. Overall the data are clearly communicated and support the conclusions. The study would be further strengthened by a few additional controls demonstrating that the cKO mice are targeting the genes as expected.
The authors show validation of the Tsc1 f/f mouse using p-56 staining as a proxy for mTOR activation. These controls are performed in Figure 6 but should be moved to an earlier figure for the Pten f/f mouse. Also, validation of the GSK3b f/f mouse doesn't appear in the paper and should be included.
For the g-ratio calculations the authors should perform statistics on the biological replicates and not the individual axons counted. Can the authors please confirm that this is the case?
Gonzalez-Fernandez and colleagues examine the role of PTEN signaling in oligodendrocyte progenitor proliferation and differentiation into mature oligodendrocytes during development and early adulthood, both in the context of a healthy CNS and during remyelination. These data are important to the myelin field as recent research has focused on the role of PI3K/Akt/mTOR signaling in myelination. The data presented here specifically point to an mTOR independent role for PTEN through GSK3β. Mice lacking PTEN specifically in OPCs exhibited continued differentiation of OPCs into mature OLs, even in early adulthood. This corresponds with increased myelin production, which is also observed in OL specific deletion of PTEN. Interestingly, loss of mTOR did not abrogate these effects of PTEN deletion, suggesting an mTOR independent mechanism. Additional data demonstrate instead that the effects of PTEN deletion require GSK3β. The data are clearly presented and make important and novel contributions. However, there are several experiments which would strengthen the conclusions.
Additionally, it is not totally clear why the TSC1 data are included, other than to suggest that TSC1 mediates effects on proliferation/differentiation via mTOR, in contrast to PTEN. However, there is a much more complete story shown for PTEN than TSC1, which fails to examine myelination, OPC proliferation, and OL survival.
1) The authors state that their Tsc1 data indicate decreased OL survival, yet do not show the data to support this statement. It is possible (as is mentioned earlier in the Results section) that the decreased mature OL number is a result of decreased proliferation and/or differentiation. Would it be possible to distinguish these possibilities by co-staining for NG2/caspase, CC1/caspase, and/or NG2/Ki67?
2) Figure 4 – The authors examine PLP and MBP by Western blotting and find that PLP is significantly increased but MBP is not. Given the different results, it would be more convincing of the structural integrity of increased myelin thickness if other myelin proteins were examined such as MOG, CNP, and MAG. Alternatively, the authors could show representative high magnification EMs showing ultrastructure.
3) The authors show that increased mature OL numbers in PTEN cKO mice correspond to increased myelin thickness and number of myelinated axons. Is there any evidence to suggest that decreased OL numbers in TSC1 cKO mice correlate with decreased myelination, either by immuostaining, Western, or EM techniques?
4) Did the mice lacking PTEN during remyelination exhibit thicker myelin or more remyelinated axons comparable to that seen during development? It is suggested in the discussion that targeting PTEN and/or GSK3B would enhance remyelination yet the increase in cell number alone does not necessarily correspond to enhanced repair.
5) With the data provided, it is difficult to assess whether GSK3B is responsible only for the differentiation enhancement that was observed in PTEN cKO mice, or if it is also responsible for myelin production by mature OLs. In the GSK3B cKO mice, would it be possible to perform either immunostaining/Western blotting for myelin proteins, or to perform EM analysis? Is it possible that PTEN mediates differentiation via GSK3B but myelin production via a different mTOR independent mechanism?
In this manuscript Gonzalez-Fernandez and colleagues report mTOR-independent effects of PTEN loss on the development of oligodendrocytes (OLs). The role of mTOR in oligodendrocyte and Schwann cell development and myelination has been thoroughly studied in the recent years, and the research so far shows that a delicate balance of mTOR activity is required at distinct steps of OL maturation. The authors focus here on whether mTOR activity is necessary for the transition of oligodendrocyte progenitors (OPCs) to myelinating oligodendrocytes (OLs) and whether PTEN deletion has an mTOR-independent effect in this process.
The authors have done a large number of experiments using conditional and inducible knockouts of Tsc1, Pten, mTOR and GSK3b. The approach taken is elegant and the findings could be important if validated. While the quantity of the data is large, the amount of analysis of each model is not sufficiently thorough. The conclusions are based on consideration of limited possibilities and some important controls are missing. Furthermore, interpretation of the results in comparison to previously published literature makes certain assumptions that are not well supported, raising doubts about the significance of the findings.
1) A recent publication from Suter lab shows the dual function of the PI3K-Akt-mTORC1 in myelination of the PNS. They suggest a molecular mechanism where mTORC1 regulates radial sorting (high mTOR activity), and mTOR and AKT act synergistically on myelin growth and myelination. The authors should cite and discuss these results in comparison to theirs briefly in the Discussion section.
2) A previous paper reported on the loss of Tsc2 gene expression in oligodendrocytes under the Olig-2 Cre promoter (DOI:10.1002/acn3.254). The authors should include this reference as well.
3) The authors use inducible Pdgfr-CreER mice with EYFP reporter to visualize the OLs. They first knockout Tsc1 to check the effect of mTORC1 hyperactivation. They find that the number of EYFP+ cells decrease significantly in CC, but not in cortex. On the other hand, EYFP+CC1+ cells (mature OLs) decrease in both areas. In cortex then, not the OPCs but pre-OL number increases, whereas in CC both increased OPC and pre-OL numbers compensate the loss in mature OL number (Figure 1K). The similar phenomenon is seen in WM of spinal cord. Can the authors comment on this distinct effect of mTOR activity on OL development in CC and cortex? Does high mTOR activity arrest OL maturation at different stages in CC and cortex?
4) Much of the data in the paper is based on the calculations of EYFP positive cells, which is taken as a reporter of Cre-recombinase activity. Another possibility is that knockout of the gene of interest in each could affect expression levels of EYFP protein. One can imagine a scenario where Cre-recombinase is active but EYFP is not detected.
5) The authors then examine whether deletion of PTEN affects the process of OL maturation and find that PTEN deletion increases the number of EYFP+ cells. In contrast to the Tsc1 deletion, they show that not in CTX but this time in CC, the number of EYFP+ OPCs do not change (Figure 2J). What could be the reason for these regional differences in Tsc1 or PTEN deletion?
6) According to the results in Figure 3, they suggest that the decrease in the density of labeled OPCs in the CC may be due to rapid differentiation of PTEN-deficient OPCs into the OL, with PTEN-harboring OPCs proliferating and replacing PTEN deficient OPCs. Have they checked this by staining the overall oligos with Olig2 and looking at the changes in the number of Olig2+/Ng2+/EYFP- cells?
7) In Figure 1, they show that Tsc1 deletion (which hyperactivates mTOR) in OPCs resulted in a decrease in the number of CC1+ cells, in Figure 6, the deletion of mTOR in OPCs had the same effect – less number of CC1+ cells. How do the authors interpret the result that Tsc1 deletion and mTOR deletion have the same effect? Moreover, how does the g-ratio and number of myelinated axons change with the deletion of mTOR alone?
8) Figure 7 is missing data: the information written in the text (subsection “OPC-targeted GSK3β inactivation enhances OL development and new myelination”) is not shown in the Figure. in vitro OPC culture – stainings with Pdgfr, O4, GalC? Was this information a result of their data or does it need a citation?
9) The effect of GSK3b on OL development and myelination is shown nicely; however, the connection between PTEN-Akt and Gsk3b is still needs to be clarified. The authors also mention this missing information in the Discussion section. In order to support this link more with the resources they have, they can check how the Gsk3b activity (phosphorylation by Akt) and the activity/level of the proteins downstream of Gsk3b change in Pten(f/f) mouse.
10) A major concern is how to reconcile the data in the current manuscript with some of the previously published results. In the Discussion section, the authors argue that non-cell specific KO of PTEN does not show the same results as theirs because of non-autonomous mechanisms. There is no data to support this interpretation at this point. Importantly, if the difference is due to non-cell autonomous effect, it is going to be extremely difficult to utilize this PTEN-Akt-GSK3b axis as a means of treatment.https://doi.org/10.7554/eLife.32021.039
- Shin H Kang
- Shin H Kang
- Shin H Kang
- Hey-Kyeong Jeong
- Young-Jin Son
- Young-Jin Son
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
We thank Dr. Seung Baek Han for the help in spinal cord demyelination surgery, and Jason Yang and Yu Mi Jeong for their excellent technical assistance. The work was supported by grants from the Ellison Medical Foundation (AG-NS-1101–13 to SHK), the National Institutes of Health (R01NS089586 to SHK and R01NS07693 to Y-JS), Shriners Hospitals for Children (85500-PHI-14 to SHK, 84298-PHI to H-KJ and 86600 to Y-JS).
Animal experimentation: All animal procedures were conducted in compliance with animal protocols (ACUP 4539 and 4568) approved by Institutional Animal Care and Committee (IACUC) at Temple University School of Medicine.
- Beth Stevens, Reviewing Editor, Boston Children's Hospital, Harvard Medical School, United States
© 2018, González-Fernández et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.