A reciprocal regulatory loop between TAZ/YAP and G-protein Gαs regulates Schwann cell proliferation and myelination

Schwann cell (SC) myelination in the peripheral nervous system is essential for motor function, and uncontrolled SC proliferation occurs in cancer. Here, we show that a dual role for Hippo effectors TAZ and YAP in SC proliferation and myelination through modulating G-protein expression and interacting with SOX10, respectively. Developmentally regulated mutagenesis indicates that TAZ/YAP are critical for SC proliferation and differentiation in a stage-dependent manner. Genome-wide occupancy mapping and transcriptome profiling reveal that nuclear TAZ/YAP promote SC proliferation by activating cell cycle regulators, while targeting critical differentiation regulators in cooperation with SOX10 for myelination. We further identify that TAZ targets and represses Gnas, encoding Gαs-protein, which opposes TAZ/YAP activities to decelerate proliferation. Gnas deletion expands SC precursor pools and blocks peripheral myelination. Thus, the Hippo/TAZ/YAP and Gαs-protein feedback circuit functions as a fulcrum balancing SC proliferation and differentiation, providing insights into molecular programming of SC lineage progression and homeostasis.

S chwann cells (SCs) produce multilamellar myelin sheaths that are essential for saltatory conduction of action potentials and axonal integrity in the vertebrate peripheral nervous system (PNS) [1][2][3] . The mutually exclusive aspects of proliferation and differentiation must be tightly regulated to allow both processes to operate properly to generate sufficient SCs for subsequent differentiation and myelination in developing peripheral nerves. Defects in SC generation and differentiation during development and regeneration may cause a failure in myelinogenesis, contributing to acquired or hereditary peripheral neuropathies associated with motor and sensory disabilities 4 . In contrast, SC over-proliferation, caused by mutations in tumour suppressor genes NF1 and NF2, results in peripheral nerve sheath tumours such as neurofibromas 5 and Schwannomas 6 , respectively.
SC development is a molecularly and ultrastructurally well-defined, multi-stage process that occurs over a protracted period of time. After specification from neural crest cells and transitioning to a SC precursor state, SC precursors and immature SCs undergo proliferation and expansion to establish a one-to-one relationship with axons during the radial sorting process. Shortly after birth, immature SCs differentiate into pro-myelinating cells and then further differentiate and form myelin sheaths around axons. SC lineage progression requires the coordinated activity of pro-myelinating factors (for example, SOX10, OCT6/POU3F1 and EGR2/KROX20) in promoting cell cycle exit and initiate differentiation, while suppressing negative regulatory factors of differentiation (for example, SOX2, NOTCH/HES and c-JUN) 7,8 . At present, the extrinsic and intrinsic signals that modulate and balance positive and negative factors to control SC proliferation and their transition to a differentiating state during peripheral myelination are not fully understood.
Recent studies indicate that TAZ/YAP are required for SC radial sorting and peripheral myelination through regulating laminin receptors 22 or KROX20 (ref. 26), and in control of myelin internodal length 27 . The nuclear localization of YAP/TAZ can be modulated by mechanical signals 22 . YAP/TAZ interact with TEAD1 to regulate expression of myelination-associated genes including Pmp22 (ref. 28), suggesting that YAP/TAZ may directly regulate the transcriptional programme necessary for SC differentiation. Given that the phenotype of Taz/Yap double mutants is much more severe than those observed in mice lacking laminin receptors 29 , it is conceivable that YAP/TAZ regulate additional targets that are responsible for the severe peripheral dysmyelinating phenotype. Currently, the direct targets regulated by YAP/TAZ during SC lineage progression have not been fully defined. Whether YAP/TAZ have a direct role in the transition from SC proliferation-to-differentiation remains unresolved.
Here we show that YAP/TAZ are expressed in SC nuclei in both culture and peripheral nerves through adulthood in mice and demonstrate that YAP/TAZ are crucial for SC proliferation in addition to myelin formation. We further map TAZ genome occupancy in SCs using chromatin immunoprecipitation and sequencing (ChIP-seq) and reveal TAZ direct targets responsible for SC proliferation and differentiation processes. We identify a TAZ target, Gnas, encoding Gas-protein, whereby Gas functions as a molecular link in a feedback regulation that attenuates TAZ/YAP activity and modulates the transition from SC proliferation to differentiation. These novel mechanistic insights into how TAZ/YAP-mediated Hippo signalling and Gas-feedback circuits control the opposing aspects of SC proliferation and differentiation provide a rationale for context-specific targeting of TAZ and YAP for treating peripheral neuropathy.

Results
Nuclear TAZ and YAP are expressed in Schwann cells. The subcellular distribution of TAZ/YAP can be modulated by various extrinsic signals such as cell density 14,16 . To examine expression and subcellular localization of TAZ and YAP in SCs, we performed co-immunolabelling with TAZ or YAP and SC markers, nuclear SOX10 and cytoplasmic S100b, in rat SCs. Both TAZ and YAP were predominantly detected in the nuclei of SCs at both low-and high-cell density (Fig. 1a). Similar nuclear localization of TAZ or YAP was also observed in SCs in the teased fibres from mouse sciatic nerves at P8 (Fig. 1b). YAP expression was abundant throughout development (Fig. 1c), and the level of the phosphorylated YAP did not exhibit substantial change at different postnatal stages ( Supplementary Fig. 1). In contrast, TAZ expression was dynamically regulated and was gradually downregulated in adulthood (Fig. 1c). In sciatic nerves, persistent nuclear TAZ/YAP localization was observed in SCs at the neonatal stages, for example, P7 (Fig. 1d) and even in fully myelinated adult stages for example, at P60 (Fig. 1e). The majority (B85%) of cells in the sciatic nerves that expressed YAP or TAZ were also SOX10 þ (Fig. 1f), suggesting an active transcriptional regulatory role of TAZ/YAP in SC development and myelination.
Taz and Yap inactivation disrupts SC development. To define the cell-autonomous role of TAZ/YAP in SC development, we bred mice carrying a Yap-floxed allele (Yap fl/fl ), a Taz-floxed allele (Taz fl/fl ) or both 30 with desert-hedgehog (Dhh)-Cre mice 31 (Fig. 2a), where Cre-mediated recombination is active in the SC lineage from embryonic day 12.5 (E12.5). These crosses generated Taz fl/fl ; Dhh-Cre mice (referred to here as Taz cKO ), Yap fl/fl ; Dhh-Cre mice (Yap cKO ) and the double mutants Yap fl/fl ;Taz fl/fl ; Dhh-Cre (TazYap dcKO ). That TAZ or YAP protein expression was efficiently abolished in SOX10 þ SCs of developing sciatic nerves was verified by immunostaining of teased nerve fibres (Fig. 2b) and western blot analysis of sciatic nerve tissues (Fig. 2c). We detected upregulation of YAP or TAZ protein expression in Taz or Yap single-mutant sciatic nerves, respectively (Fig. 2c), suggesting a compensatory effect in each other's absence. Both Taz cKO and Yap cKO mice appeared phenotypically normal as compared to heterozygous control littermates at P60. At P7, sciatic nerves in these mutants were thick and opaque, and comparable to wild-type controls (Fig. 2d). Ultrastructural analysis of sciatic nerves from Taz cKO and Yap cKO mice had the expected 1:1 ratio of SC to axons and showed no significant impairments in myelination (Fig. 2d,e). Total SOX10 þ SCs and EGR2 þ myelinating SCs in Taz cKO and Yap cKO mice were unaltered relative to wild-type controls ( Fig. 2f-h). The absence of apparent myelination defects in Taz or Yap single-mutants suggests that TAZ and YAP function redundantly during SC development.
We then analysed the double mutant TazYap dcKO and Taz cKO heterozygous for Yap (Taz fl/fl ; Yap fl/ þ ; Dhh-Cre, referred to as Yap cKO Taz Het ) and Yap cKO heterozygous for Taz (Yap fl/fl ; Taz fl/ þ ; Dhh-Cre, referred to as Taz cKO Yap Het ) to investigate dosage-dependent effects of TAZ/YAP on SC development. The Yap cKO Taz Het mice were phenotypically indistinguishable from control littermates with properly sorted and myelinated large calibre axons at P7 (Fig. 2d,e). In contrast, Taz cKO Yap Het sciatic nerves appeared thin and translucent and had radial sorting and myelination deficiencies (Fig. 2d,e and Supplementary Fig. 2). In addition, SOX10 þ SCs and EGR2 þ SCs were reduced in number in Taz cKO Yap Het mice ( Fig. 2f- Supplementary Fig. 2). The majority of TazYap dcKO SCs failed to develop beyond the immature stage, as revealed by the absence of myelinating profiles (Fig. 2d,e), indicating that Taz/Yap are required for SC lineage progression and subsequent myelination.
TAZ and YAP are critical for SC proliferation. SCs must proliferate to match axon number for radial sorting. The severe reduction in SC number in TazYap dcKO sciatic nerves prompted us to hypothesize that the arrest of SC lineage progression when TAZ/ YAP are lost is a consequence of defects in SC proliferation and expansion. We quantified SC proliferation in sciatic nerves in SOX10 þ SCs by Ki67 immunolabelling and bromodeoxyuridine (BrdU) incorporation at P1 and P7, and observed a stark reduction in Ki67-expressing and BrdU-labelling SCs in TazYap dcKO mice ( Fig. 3a-c). We did not detect significant differences in cell death in the TazYap dcKO mice ( Supplementary Fig. 3   using siRNA in purified rat SCs under growth conditions. Reduced expression of Taz-or Yap-decreased SC proliferation as assayed by BrdU incorporation (Fig. 3d,e). Inhibition of both Taz and Yap resulted in more severely compromised SC proliferation than when expression of a single factor was inhibited (Fig. 3d,e). Conversely, expression of constitutively active form of TAZ (Taz 4SA with serine-to-alanine mutations of four phosphorylation sites) 32 , or YAP (Yap S112A with mutation of the phosphorylation site serine 112 to alanine) 33 , robustly enhanced SC proliferation (Fig. 3f,g). Together, our data indicate that activation of TAZ/YAP promotes SC proliferation, consistent with their pro-growth and oncogenic function in other contexts 34 .
TAZ/YAP control the initiation of SC differentiation. To determine the role of TAZ/YAP in the differentiation process, we inactivated Taz/Yap in immature SCs in neonates by using a tamoxifen-inducible SC-expressing Plp-CreERT driver 35 carrying a floxed CAG-GFP (green fluorescence protein) reporter (ccGFP) 36 to bypass the possible impact of the Taz/Yap loss on early SC development. Mice were treated with tamoxifen from P0 to P9 to induce deletion of Taz and Yap; we refer to these mice as TazYap iKO mice (Fig. 4a). TAZ/YAP were efficiently depleted in GFP reporter-expressing cells after tamoxifen treatment (Fig. 4b).
We detected a severe reduction in the percentage of EGR2 þ mature SCs while an increase in the percentage of the immature OCT6 þ SCs with Taz/Yap deletion ( Fig. 4c-f). Compared to control nerves at P14, about 42% of the large axons were not properly sorted and remained unmyelinated in TazYap iKO nerves ( Fig. 4g-i), suggesting that YAP/TAZ are required for proper initiation of SC differentiation from immature SCs and their subsequent myelination. Intriguingly, tamoxifen-induced deletion of Taz/Yap in adult mice did not substantially alter myelin sheath thickness nor the integrity in peripheral nerves ( Supplementary  Fig. 4), despite the fact that these mice exhibited severe tremors, ataxia and finally mortality within one month after tamoxifen-  induced Taz and Yap deletion. These observations suggest a crucial role for TAZ/YAP in initiating SC differentiation. There appears to be no immediate requirement for YAP/TAZ for peripheral myelination maintenance, however, due to lethality we were unable to examine mice more than a month post-ablation.
TAZ/YAP regulate SC proliferation and myelination programs. Next, we sought to define the underlying mechanisms whereby TAZ/YAP regulate SC growth and differentiation by comparing the transcriptome profiles of P5 sciatic nerves consisting of proliferating and differentiating SCs from control, Taz cKO Yap Het , and TazYap dcKO mice. Consistent with the intermediate dysmyelinating phenotype of Taz cKO Yap Het mice, the changes in the expression pattern were intermediate between control and TazYap dcKO mice. We therefore focused our comparison on the transcriptomes of control and TazYap dcKO mice. Gene ontology analysis revealed that the downregulated genes in TazYap dcKO mice were enriched for cell proliferation regulation, lipid biosynthesis and myelination (Fig. 5a,b), consistent with deficient SC proliferation and the dysmyelination phenotype in these mutants.
Among the differentially expressed genes, we identified a set of Hippo pathway signatures related to cell growth, including Amotl2, Fgf1, Ect2 and Ddah1 (refs 37,38; Fig. 5b), that were downregulated in TazYap dcKO , consonant with TAZ/YAP as the key effectors in Hippo signalling. In addition, expression of critical myelination regulators such as Egr2 and Zeb2/Sip1, as well as myelin genes, Mbp, Pmp22 and Erbb2/3, were dramatically downregulated in TazYap dcKO nerves, with a concomitant elevation of myelination inhibitory genes such as Hes1 and Egr1  Table 1). In accordance with the reduction of SC number, we detected a decrease in a set of cell cycle-related genes Ccnj, Cdk6, Pim3 and Mycn (Fig. 5b). Consistent with severe radial sorting defect in the TazYap dcKO nerves, expression of genes encoding critical regulators of radial sorting including Dag1, Itgb1, Itga6 and Itga8 (ref. 22) was reduced (Fig. 5b). In agreement, analysis of the global gene regulatory network also showed dysregulation of genes involved in cell proliferation, mitotic cell cycle and in myelination in TazYap dcKO nerves (Fig. 5c). Since the apparent gene expression changes might reflect a lower proportion of SCs in TazYap dcKO nerves, to investigate whether TAZ/YAP regulate gene expression programs of SC proliferation and differentiation in a cellautonomous manner, we carried out siRNA-mediated knockdown of Taz and Yap in rat SCs. Expression of Hippodownstream targets (Fig. 5d) and cell cycle/proliferation-related genes (Fig. 5e), as well as SC differentiation-associated genes ( Fig. 5f) was significantly reduced in SCs that underwent Taz and Yap knockdown. Collectively, these data suggest that TAZ/YAP control regulatory networks for SC proliferation and differentiation.
TAZ cooperates with SOX10 to regulate SC differentiation. Given that TAZ/YAP have similar genomic binding profiles 39 , to gain insights into TAZ/YAP target genes, we performed ChIP-seq analysis in primary rat SCs to assess the chromatin occupancy of TAZ, which has a more prominent role than YAP in SC development. We analysed the distribution of TAZ-binding regions and found that a large proportion of TAZ-binding peaks (B73%) mapped to the evolutionarily conserved enhancer elements in promoter and intergenic regions (Fig. 6a), consistent with previous reports in other contexts 39 . These peaks were associated with an activating histone mark H3K27Ac (Fig. 6b), and a bimodal distribution of H3K27Ac around TAZ peaks was observed (Fig. 6c). Gene ontology revealed that these TAZ direct targets are associated with biological processes related to cell proliferation, as well as glial/SC differentiation (Fig. 6d). TAZ occupancy was detected around the promoter and enhancer regions of cell growth-related canonical Hippo pathway genes including Amotl2, Fgf1 and Ddah1 as well as cell cycle regulators including Cdk6 (Fig. 6e). Accompanying the activating H3K27ac mark, TAZ occupancy likely activates expression of these cell proliferation-related genes, consistent with downregulation of their expression in TazYap knockout nerves or knockdown SCs (Fig. 5).
To further explore the potential co-regulators of TAZ, we analysed the potential transcriptional co-regulator binding motifs enriched in TAZ-targeted sites using HOMER motif-discovery algorithm 40 . Motif analysis revealed a strong enrichment of Regulation of lipid biosythetic process the known TEAD-binding consensus motif GGAATG 41 within TAZ-binding sites (Fig. 6f), suggesting that TAZ exerts its transcriptional control via TEAD, consistent with a previous study of YAP genomic occupancy 41 .
In addition to the TEAD-consensus motif, we also found that B34% of TAZ-binding sites were enriched with the Sox-consensus binding motif C(T/A)TTG(T/A)(T/A) and matched most closely to that of SOX10 (ref. 42), a critical factor for SC lineage progression 3 . Further, there was enrichment for the AP-1 (c-Jun/c-Fos complex) motif TGACTCA, consistent with a previous report in other cellular contexts 39 (Fig. 6f). This suggests that TAZ/YAP/TEAD cooperate with distinct co-regulators to modulate the gene programs necessary for SC lineage progression.
Analysis of TAZ-binding elements together with SOX10 binding sites in rat sciatic nerves 42 revealed that TAZ co-occupation with SOX10 at the H3K27ac-marked promoters/enhancers of a set of myelination-associated genes including Mbp, Mpz, Pmp22 and Mal, as well as SC differentiation regulators Sox10, Erbb2/3 and Zeb2 (Fig. 6g-i). Consistently, TAZ and SOX10 overexpression in combination enhanced the activity of reporter genes driven by the promoters of myelin genes Mbp and Mpz, which are targeted by SOX10 and TAZ (Fig. 6j), while suppressing those of SC differentiation inhibitors Hes1, Hes5 and c-Jun (Fig. 6k), suggesting that TAZ cooperates with SOX10 to promote the transcriptional programme required for SC differentiation.
TAZ regulates G-protein signalling. Combinatorial analysis of transcriptome profiles and TAZ genome-wide occupancy revealed a set of G-protein signalling genes that were TAZ-occupied and dysregulated in Taz/Yap-deficient SCs (Fig. 7a). Consistent with their role as agonists of TAZ/YAP activation 43 , the genes encoding Ga proteins including Gnai1, Gnaq, Gnao1 and Gna11 were downregulated in TazYap dcKO and Taz cKO Yap Het nerves (Fig. 7a). In contrast, Gnas, which encodes Gas, was upregulated in Taz/Yap mutant nerves (Fig. 7a). Consistently, siRNA knockdown of Taz/Yap in SCs resulted in an upregulation of Gnas, while inhibiting expression of Gnai1 and Gnaq (Fig. 7b).
TAZ and SOX10 co-occupancy was detected in the enhancer region of Gnai1 in association with H3K27Ac, suggesting that TAZ and SOX10 co-targeting enhances expression of Gnai1, which encodes an inhibitory Gai-protein 19 (Fig. 7c). In contrast, TAZ targeted to the Gnas gene locus was not associated with enrichment of H3K27Ac or SOX10 (Fig. 7d). This prompted us to hypothesize that TAZ represses Gnas transcription. Indeed, constitutive expression of TAZ or YAP in rat SCs suppressed Gnas expression (Fig. 7e). The enrichment of TAZ occupancy on the targeted elements of Gnas and Gnai1 loci was further confirmed ChIP-qPCR in rat SCs (Fig. 7f). These observations suggest that TAZ directly targets and surprisingly inhibits Gnas expression, while activating Gnai1 expression, indicating a context-specific TAZ activity.
We detected an increase in Gnas-encoded Gas protein expression in SCs cultured under differentiation compared to proliferation conditions (Fig. 7g). Similarly, expression of Gas protein was detected at the neonatal stage P0, when immature SCs undergo proliferation, and peaked at the perinatal stage P7, the onset of SC myelination, but then subsided in late postnatal stages and adulthood (Fig. 7h). To examine the function of Gnas in SC development in vivo, we specifically ablated Gnas in SCs using Dhh-Cre (Fig. 7i). Gnas cKO mice were born at a Mendelian frequency but died around postnatal week 2. Gnas cKO mice exhibited severely impaired movement characterized by paralyzed hindlimbs and tremors, similar to the phenotype of TazYap dcKO mice. In contrast to controls, Gnas cKO sciatic nerves appeared thin and translucent (Fig. 7j), indicative of a myelination deficiency. At P0, when SCs had started to establish a 1:1 relationship with axons in control nerves, a majority of SCs in Gnas cKO mice remained associated with multiple large calibre axons (Fig. 7k), suggesting incomplete radial sorting. At P10, control SCs had advanced to myelinate segregated large diameter axons; however, in Gnas cKO mice, SCs were essentially arrested at the pro-myelinating stage and failed to initiate myelin sheaths (Fig. 7k). Large unsorted axon bundles were also frequently detected in Gnas cKO sciatic nerves ( Fig. 7k; Supplementary  Fig. 5). The severe myelinogenesis defect in Gnas cKO mice was corroborated by the absence of myelin proteins such as MBP and MAG in Gnas-mutant nerves (Fig. 7l).
As TAZ directly targets the Gnas locus and inhibits its expression, the observation that Gnas cKO nerves exhibited a similar morphological phenotype to TazYap dcKO nerves appeared to be counterintuitive since both Gnas and Taz/Yap are needed for SC differentiation. We therefore analysed gene expression profiles of Gnas cKO sciatic nerves. Strikingly, gene ontology analysis reveals that the genes upregulated Gnas cKO nerves compared to controls at P5 are involved in cell cycle regulation, cell division and DNA replication; this is the reverse of differential gene-set enrichment when compared with that of TazYap dcKO nerves (Fig. 7m).

Gas signalling antagonizes YAP/TAZ-mediated SC proliferation.
In Gnas cKO sciatic nerves at P7, we detected an increase in OCT6 þ immature/pre-myelinating SCs, which failed to mature into EGR2 þ myelinating SCs (Fig. 8a,b). Gnas cKO SCs persistently expressed the immature marker SOX2, suggesting a blockade of SC lineage progression beyond the immature stage (Fig. 8c, d). Notably, whereas control SCs gradually stop proliferating as myelination proceeds, Gnas cKO SCs remained robustly proliferative, as revealed by an increase in Ki67 þ SCs (Fig. 8e,f), resulting in an escalation in immature SC number in the mutant nerves. These data suggest that SCs fail to exit the cell cycle in the absence of Gnas.
To further confirm that downregulation of Gnas by TAZ promotes SC proliferation, we depleted cultured SCs of Taz or Gnas or both by siRNA-mediated inhibition in proliferation media without forskolin to prevent upstream Gas-protein signalling. TAZ deficiency efficiently reduced the numbers of BrdU þ proliferative SCs, consistent with its critical role in SC proliferation, while inhibition of Gnas expression alone in SCs recapitulated the hyper-proliferation phenotype seen in postnatal Gnas cKO peripheral nerves (Fig. 8g-i). Strikingly, inhibition of both Taz and Gnas in SCs restored proliferation to levels in cells treated with control siRNAs (Fig. 8g-i). Conversely, we performed Gnas gain-of-function studies in SCs maintained in mitogen-depleted media. Despite the absence of neuregulin and forskolin, conditions in which SC proliferation was severely limited (Fig. 8j), constitutive TAZ activation forced a population of SCs to proliferate, as indicated by an increase in BrdU þ and Ki67 þ SCs, which were barely detected in control SCs (Fig. 8j-l). Combined expression of activated TAZ and GsCA, a constitutively activated form of Gas (ref. 44), effectively rescued the over-proliferation defect observed in TAZ-expressing SCs (Fig. 8j-l). Therefore, the balance between TAZ/YAP and Gas activities is critical for regulating SC proliferation and subsequent maturation.

Discussion
The opposing cell proliferation and differentiation processes are tightly regulated to ensure proper timing of SC lineage progression. The underlying mechanisms that coordinate these processes are not fully understood. Our studies with SC lineagespecific and developmentally regulated mutagenesis approaches indicate that Hippo signalling effectors TAZ/YAP are required for SC proliferation, radial sorting and differentiation. SC proliferation is a prerequisite for the subsequent events of differentiation and myelination. Activation of TAZ/YAP is sufficient to induce SC proliferation, consistent with genome occupancy analysis showing that TAZ targets cell proliferation-associated genes including canonical Hippo-downstream effectors and cell cycle regulators.
A recent study suggested that dysmyelination in mutants in which Taz/Yap were deleted using the P0-Cre line are mainly due to radial sorting defects 22 . In the present study, we found that mice with Taz/Yap deletion in SC precursors using Dhh-Cre exhibit severe deficits in SC number and proliferation. The discrepancy could be due to the Cre lines: Dhh-Cre is expressed in earlier SC precursors beginning at E12.5 than is P0-Cre, which is expressed around E13. 5 (refs 31,45), raising the possibility that SC proliferation is regulated during the window between the inception of Dhh and P0 promoter activities. Similarly, the lack of substantial radial sorting defect in Taz cKO single mutants at P7 (Supplementary Fig. 2) may attribute to Dhh-Cre-mediated recombination, although we cannot exclude the possibility that there will be a sorting defect at later postnatal stages as examined in the previous report 22 . Our analysis demonstrated that the TAZ/YAP effect is dosage-dependent since a heterozygous allele in either Yap or Taz compensates for SC proliferation defects to certain extent. Deletion of Yap or Taz resulted in compensatory upregulation of expression of the non-deleted Taz or Yap gene product (Fig. 2c), suggesting that YAP and TAZ are functionally redundant and that their expression levels are tightly regulated to maintain homeostatic state of SC development.
Although myelination-associated laminin receptor genes and Krox20 are reported as YAP/TAZ targets 22,26 , our genome occupancy and transcriptome profiling analyses revealed a global view of TAZ-targeted gene loci including those encoding essential SC differentiation regulators SOX10 and ZEB2 along with neuregulin receptors ERBB2/3 (refs 46,47), in addition to the reported targets. Our data further indicate that TAZ co-occupies with SOX10 on TEAD-binding sites in SCs, consistent with a recent finding of association of SOX10 binding sites with TEAD in the PNS, but not the CNS 48,49 . TAZ/SOX10 co-targeted elements are also associated with the activating H3K27Ac in the gene loci including Sox10, Zeb2 and Erbb2/3, encoding factors that positively regulate SC differentiation and myelination [50][51][52][53] . Surprisingly for what has been hitherto known as a co-activator, TAZ occupancy without SOX10 and H3K27Ac marks appears to repress expression of the genes associated with negative regulation of SC differentiation, suggesting that TAZ functions as a repressor in this context. Furthermore, by using Plp-CreERTinduced Taz/Yap deletion, we showed that Taz/Yap are required for initiation of differentiation from immature SCs and myelin formation. Intriguingly, similar to a recent study from Grove et al. 26 , deletion of Taz/Yap in adult mice by Plp-CreERT resulted in severe tremors and ataxia. The mice died within 1 month after tamoxifen treatment. In contrast to the Grove et al. study, we did not observe appreciable alteration in myelin thickness or axonal integrity in peripheral nerves after 36 days post tamoxifen treatment ( Supplementary Fig. 4). Although the exact mechanism underlying the discrepancy remains to be determined, it is possible that the use of different Plp-CreERT lines 35,54 or genetic backgrounds may account for different phenotypes in myelin maintenance in sciatic nerves. In addition, TAZ/YAP may regulate physiological properties of SCs and axo-glial interactions, or perhaps, that the essential function of TAZ/YAP in Plp-expressing non-SCs such as spleen, liver, kidney and lung 55 may cause animal death besides demyelination. YAP was reported to play a role in oligodendrocyte morphology and maturation in response to mechanical factors 56 , however, we did not detect overt myelin abnormalities in the adult corpus callosum or spinal cord of TazYap iKO mice ( Supplementary Fig. 6), suggesting that TAZ/YAP are not required for CNS myelin maintenance. Collectively, our observations suggest that TAZ/YAP are critical for SC proliferation and the myelination onset during SC development.
A recent study indicates that TAZ and YAP are present predominantly in the cytoplasm of SCs at high cell density, and that mechano-stimulation promotes their translocation from the cytoplasm to the nucleus 22 . Intriguingly, in cultured SC and in intact sciatic nerves, we detected expression of TAZ and YAP primarily in the nuclei of SCs at both low and high density, consistent with their function as transcriptional regulators. The underlying discrepancy is likely due to different culture conditions since our finding is consistent with the previous study 22 when SCs were plated in glass coverslips, indicating that intracellular localization is context-dependent for TAZ and YAP.
Recent studies indicate that Hippo effectors acts downstream of GPCR signalling 19 . Activation of Gas-coupled GPCRs has been shown to inhibit TAZ/YAP nuclear translocation and their activity 19,57 . Consistent with Gas function, Gpr126, which is able to couple to Gas, is required for proper SC development and myelination 58,59 . Strikingly, we found that Gas-encoding Gnas itself is a direct transcriptional target of TAZ. Despite TAZ being observed thus far to be a co-activator, activation of TAZ/YAP inhibits expression of Gnas. Gas protein levels are low under proliferative conditions and upregulated during SC differentiation. Consistently, Gnas deletion leads to excessive SC proliferation while inhibiting SC differentiation, suggesting that Gas protein levels regulate the transition of SC proliferation to differentiation. Radial sorting defects in Gnas mutants appear less severe compared to the Taz/Yap double mutant 26,58 ( Supplementary Figs 2 and 5) or Gpr126 mutant 22,26,58 , this suggests that these pathways may regulate radial sorting via additional mechanisms other than Gas-dependent mechanisms. Nonetheless, our findings identify a crucial role of the reciprocal signalling of TAZ/YAP and Gas in regulating SC proliferation-todifferentiation transition. Gnas knockdown in SCs lead to a downregulation of cAMP levels ( Supplementary Fig. 7). Intriguingly, Oct6 appears upregulated in Gnas and Taz/Yap mutant sciatic nerves 26 . Since Oct6 is a downstream target of cAMP signalling in SCs, it is possible that Gas functions in cAMP-dependent and -independent mechanisms during SC development in sciatic nerves, or perhaps, other pathways may activate Oct6 expression in the absence of Gas-signalling in Gnas-mutant nerves.
Given that YAP/TAZ remain active in adult SCs, Gas downregulation at late developmental stages in differentiated SCs may allow the pro-differentiation effects of TAZ/YAP, in collaboration with SOX10, on regulation of the SC myelination programme to prevail (Fig. 8m). This suggests that changes in G-protein levels may account for a switch in the net effect of YAP/TAZ on proliferation versus differentiation during SC development in a stage-dependent manner.
Our ChIP-seq data indicate that TAZ targets an intronic region in the Gnas locus. Although transcriptional regulators can bind to regions of open chromatin without functional effect 60 . ChIP-PCR confirms that TAZ binds to the targeted sites on the Gnas locus. By overexpression of TAZ or YAP, Gnas transcription is attenuated, while reduction of TAZ/YAP by siRNA knockdown leads to an upregulation of Gnas, suggesting that TAZ may transcriptionally represses Gnas expression. It is worth noting that since the expression pattern of Gnas does not strictly correlate with that of Taz and Yap mRNA, it is unlikely that TAZ is the sole repressor of Gnas expression, which could be also regulated by other mechanisms. Conversely, Gnas deficiency leads to an upregulation of Taz/Yap as well as Gli1-3 expression in Gnas-mutant nerves ( Supplementary Fig. 8). This is consistent with Gas function as a tumour suppressor that inhibits Sonic hedgehog signalling and Hippo signalling-mediated cell proliferation in medulloblastoma or basal-cell carcinoma 44,61,62 , suggesting a reciprocal and antagonistic function between Gas signalling and Hippo/hedgehog pathways.
Our findings place Gas-mediated signalling in an unanticipated role as a key modulator that antagonizes TAZ/YAP effects during the onset of SC differentiation. We propose that the TAZ/YAP/Gas regulatory loop acts as a feedback circuit that balances SC proliferation and differentiation (Fig. 8m). As constitutive activation of either YAP or TAZ enhances proliferation of cultured SCs, this Gas regulatory loop may serve to limit SC proliferation. By contrast, Gas activation results in reduced SC proliferation, whereas Gnas deletion leads to a drastic increase in SC proliferation, consistent with a tumour suppressor function of Gnas in other contexts 44,61 . Thus, our findings establish Gas as a critical node in the inhibitory feedback-loop that antagonizes TAZ/YAP activity in regulating the balance between SC proliferation and differentiation. Future studies aimed at modulating or fine-tuning the dynamic balance between YAP/TAZ and Gas may lead to therapeutic strategies to promote myelination in the PNS while inhibiting SC over-proliferation that occurs during formation of peripheral nerve sheath tumours such as neurofibromas and Schwannomas.

Methods
Animals. Mice homozygous for floxed alleles of Taz (Taz lox/lox ); Yap (Yap lox/lox ) 30 and Gnas lox/lox mice 63 were crossed with mice carrying Dhh-Cre 31 to generate Taz and Yap single and double mutant mice or Gnas cKO mice, respectively. Inducible knockouts, iKO, for recombination perinatally or following sciatic nerve transection were generated by crossing Taz lox/lox ;Yap lox/lox mice with the inducible Cre recombinase Cre-ERT2 under the control of the Plp promoter (Plp-Cre-ERT) 35 followed by tamoxifen injection. Taz lox/lox ;Yap lox/lox ; Plp-Cre-ERT mice were also bred to ccGFP reporter mice 36 . Animals of either sex were used in the study, and littermates were used as controls unless otherwise indicated. The mouse strains used in this study were generated and maintained on a mixed C57Bl/6;129Sv background and housed in a vivarium with a 12-hour light/dark cycle. No more than 4 adult mice are housed in the same cage per IACUC regulations at CCHMC. All animal use and studies were approved by the Institutional Animal Care and Use Committee at Cincinnati Children's Hospital, USA.
Tamoxifen induction of gene deletion. Tamoxifen (Sigma) was dissolved to a stock concentration of 20 mg ml À 1 in a vehicle of ethanol and sunflower seed oil (1:9 v/v). For perinatal tamoxifen injections, tamoxifen stock was injected i.p. at 2 mg per 100 ml for 10 consecutive days to lactating mothers, thus administering tamoxifen to pups, beginning at P0. Sciatic nerves of pups were analysed on P14 for immunofluorescence staining and electron microscopy. For adult tamoxifen treatment, 100 ml (2 mg per 100 ml) was administered by i.p. injection once daily for 5 consecutive days to 6-8 week-old mice. Mice were treated again for 5 days after a 1-day rest period. Control mice were treated identically.
Primary Rat SC culture. Rat SCs from sciatic nerves of newborn rats (1-2 days-old) were isolated as described previously 64 . SCs were grown routinely in DMEM/10% fetal bovine serum (FBS) (Life Technologies), supplemented with 10 ng ml À 1 Neuregulin 1 (R&D Systems), and 5 mM forskolin (Sigma), plus L-glutamine and penicillin/streptomycin, hereafter denoted as SC growth/ proliferation medium. Cells between passages 2 and 6 were used in all experiments. More than 95% SC purity was achieved, assessed by positive SOX10 and S100b immunoreactivity. To initiate differentiation, SCs were cultured in differentiation medium containing DMEM/0.5% FBS and 1 mM dibutyl cyclic AMP (Sigma) with L-glutamine and penicillin/streptomycin, for the length of time indicated in the text, depending on the assays used. All tissue culture containers and glass coverslips (Carolina Biological cat# 633029) were coated with 50 mg ml À 1 poly-L-lysine (Sigma) in phosphate buffered saline (PBS) for at least 30 min at room temperature and then rinsed with distilled water. For BrdU incorporation analysis, BrdU at 20 mM was added to purified rat SCs seeded on coverslips 24 h or otherwise indicated, prior to fixation in 4% (w/v) paraformaldehyde (PFA).
Immunofluorescence staining. The sciatic nerves of mice at defined ages were dissected and fixed for 15 min in 4% PFA in 0.1 M sodium phosphate buffer (pH 7.4), embedded in OCT, cryoprotected in 25% sucrose and sectioned at 12 mm as longitudinal or cross-sections using a cryostat or at 30 mm using a vibratome. Teased fibres were prepared by separating and teasing nerve bundles into individual fibres with acupuncture needles in a drop of PBS (pH 7.4) on glass slides followed by air-drying.
For BrdU pulse labelling, animals were injected subcutaneously or intraperitoneally with 100 mg BrdU kg À 1 body weight 2 h prior to sciatic nerve collection. Tissue sections or cells were permeabilized and blocked in blocking buffer (0.3% Triton X-100 and 5% normal donkey serum in PBS) for 1 h at 25°C, followed by incubating with primary antibodies overnight at 4°C. We Cell Signaling Technology, #9661, dilution 1:500), S100b (Mouse; Sigma, SAB1402349, dilution 1:1,000). Secondary antibodies conjugated to Cy2, Cy3 or Cy5 were from Jackson ImmunoResearch Laboratories (dilution of secondary antibodies is 1:1,000). All images were acquired using a Nikon C2 þ confocal microscope.
Electron microscopy and morphometric analysis. Mice were perfused with 4% PFA, 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.2. Sciatic nerves were dissected and fixed in the same fixative solution overnight. Nerves were rinsed in PBS, postfixed in 1% OsO4 in PBS for 1 h, dehydrated in graded ethanol, infiltrated with propylene oxide, and embedded in Epon. Semithin sections were stained with toluidine blue, and ultrathin sections were stained with lead citrate. The morphometric measurements of axonal sorting defects were performed using electron micrographs of ultrathin sections and analysed using NIH Image J software (http://rsb.info.nih.gov/ij/). An entire sciatic nerve cross-section per animal was reconstructed by merging several high magnification photographs taken by bright field microscopy at 100 Â magnification. The number of myelinated axons per nerve were analysed in ultrathin sections using a JEOL 1200 EXII electron microscope.
Transient transfections and luciferase assays. For plasmid transfections, HEK293 cells in DMEM/10% FBS were transfected with expression vectors using Polyjet (SignaGen, SL100688) per the manufacturer's protocol and assayed for immunocytochemistry or quantitative polymerase chain reaction with reverse transcription (qRT-PCR) analysis.
For siRNA knockdown in SCs, we used Lipofectamine RNAiMAX (Life Technologies, cat# 1377850) per manufacturer's instructions. SCs were plated in SC proliferation medium. Prior to siRNA knockdown, SC medium was changed to DMEM/10% FBS with 10 ng ml À 1 Neuregulin 1 only. SCs were transfected with scrambled siRNA control or specific siRNAs against Taz, Yap or Gnas for 48 h before being subject to immunocytochemistry or qRT-PCR analysis. For double siRNA knockdown, two gene specific siRNAs were prepared in the same transfection mix before adding to the cells. siRNAs were purchased from Sigma-Aldrich with the following catalogue numbers: for rat Gnas, SASI_Rn02_ 00202731 and SASI_Rn02_00202732; for rat Taz, SASI_Rn02_00202747 and SASI_Rn01_00120445; for rat Yap, SASI_Rn01_00114056 and SASI_Rn01_ 00114054.
Proliferation was assessed by immunostaining for Ki67 and BrdU in SOX10 þ SCs seeded on poly-L-lysine (PLL)-coated coverslips. Differentiation was assessed by immunostaining for OCT6 or EGR2. Multiple images were taken from each coverslip to obtain representative images from all areas of the coverslip, and at least 400 cells/coverslip were counted using Nikon NIS-Elements software.
For reporter assays, HEK293 cells were transfected with pGL3-luciferase reporters driven by indicated promoters to TSS sites (Mbp, À 1,323 bp to þ 3 bp; Mpz, À 1,100 kb to þ 0 bp; Hes1, À 467 bp to þ 46 bp; Hes5; À 800 bp to þ 73 bp; c-Jun, À 225 bp to þ 150 bp) using Polyjet per the manufacturer's protocol and assayed 48 h, post-transfection for luciferase activity by using a Promega luciferase assay kit according to the manufacturer's instructions. The pSV-b-galactosidase control vector was included to control for variable transfection efficiencies between independent experiments.
RNA isolation and quantitative real-time PCR. RNA from purified rat SCs or mouse sciatic nerves was extracted using TRIZOL (Life Technologies). cDNA was synthesized from RNAs using iScript Reverse Transcription Supermix (BioRad) according to the manufacturer's instructions. qRT-PCR was performed using the StepOnePlus Real-time PCR System (Applied Biosystems). qRT-PCR was performed using quantitative SYBR green PCR mix (BioRad). Comparative DDCT were performed and normalized to Gapdh as an internal control. PCR primer sequences are available in Supplementary Table 2.
Enzyme-linked immunoassay for cAMP. Rat SCs were plated on 12-well plates and cultured for 1 day. Cells were transfected with control and Gnas siRNA and cultured in the SC growth medium without forskolin overnight. The cAMP level in SC lysates was assayed following the manufacturer's protocol (Cell Signaling, cat# 4339).
ChIP-seq peak-calling and data analysis. All sequencing data were mapped to rat genome assembly rn5 and peak calling was performed using model-based analysis of ChIP-seq version 1.4.2 (http://liulab.dfci.harvard.edu/MACS) with default parameters, to get primary binding regions. The ChIP-seq were done in two biological replicates. The numbers of reads per replicate were B20-21 million for TAZ ChIP-seq, 19-20 million for H3K27Ac ChIP-seq, respectively. To ensure that our data were of high quality and reproducibility, we called peaks with enrichment Z10-fold over control (Pr10 À 9 ) and compared the peak sets using the ENCODE overlap rules. These identified primary regions were further filtered using the following criteria, to define a more stringent protein-DNA interactome: (1) the P value cut-off was set to r10 À 9 ; (2) an enrichment of 6.8-fold and peak height 45.
The genome-wide distribution of protein binding regions was determined by HOMER (http://homer.salk.edu/homer/index.html) in reference to Ensembl RGSC3.4.61 release. Motif enrichment analysis was performed using Homer software. For all ChIP-seq data sets, WIG files were generated with model-based analysis of ChIP-seq, which were subsequently visualized using Mochiview v1.46. Taz and H3K27Ac ChIP-seq heatmaps were ordered by strength of binding. The heatmaps were drawn using the Heatmap tools provided by Cistrome (http://cistrome.org/ap).
Statistical analysis. All analyses were done using Microsoft Excel or GraphPad Prism 6.00 (San Diego, California, www.graphpad.com). Data are shown in dot plots or bar graphs as mean ± s.e.m., Po0.05 is deemed statistically significant. Data distribution was assumed to be normal, but this was not formally tested. Count data was assumed to be non-parametric, and appropriate statistical tests were used. Statistical analysis was performed by two-tailed unpaired Student's t-tests, Mann-Whitney test, one-way analysis of variance (ANOVA) with post hoc analysis by Tukey's multiple-comparison test, or as indicated. Quantifications were performed from at least three experimental groups in a blinded fashion. No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those generally employed in the field. No randomization was used to collect all the data, but they were quantified blindly. No animals or data points were excluded from analyses.
Data availability. All the RNA-seq and ChIP-seq data have been deposited in the NCBI Gene Expression Omnibus (GEO) under accession number GSE94990.