1. Neuroscience
Download icon

Axon-dependent expression of YAP/TAZ mediates Schwann cell remyelination but not proliferation after nerve injury

  1. Matthew Grove
  2. Hyunkyoung Lee
  3. Huaqing Zhao
  4. Young-Jin Son  Is a corresponding author
  1. Shriners Hospitals Pediatric Research Center and Center for Neural Repair and Rehabilitation, Temple University, United States
  2. Department of Anatomy and Cell Biology, Temple University, United States
  3. Department of Clinical Sciences, Lewis Katz School of Medicine, Temple University, United States
Research Advance
  • Cited 1
  • Views 700
  • Annotations
Cite this article as: eLife 2020;9:e50138 doi: 10.7554/eLife.50138

Abstract

Previously we showed that YAP/TAZ promote not only proliferation but also differentiation of immature Schwann cells (SCs), thereby forming and maintaining the myelin sheath around peripheral axons (Grove et al., 2017). Here we show that YAP/TAZ are required for mature SCs to restore peripheral myelination, but not to proliferate, after nerve injury. We find that YAP/TAZ dramatically disappear from SCs of adult mice concurrent with axon degeneration after nerve injury. They reappear in SCs only if axons regenerate. YAP/TAZ ablation does not impair SC proliferation or transdifferentiation into growth promoting repair SCs. SCs lacking YAP/TAZ, however, fail to upregulate myelin-associated genes and completely fail to remyelinate regenerated axons. We also show that both YAP and TAZ are redundantly required for optimal remyelination. These findings suggest that axons regulate transcriptional activity of YAP/TAZ in adult SCs and that YAP/TAZ are essential for functional regeneration of peripheral nerve.

Introduction

YAP (Yes-associated protein) and TAZ (Transcriptional coactivator with PDZ-binding motif), are paralogous transcription coactivators, chiefly known as potent stimulators of cellular proliferation in diverse developing (von Gise et al., 2012; Zhang et al., 2012; Xin et al., 2013; Cotton et al., 2017) and neoplastic (Yu et al., 2015; Zanconato et al., 2016; Moon et al., 2018) tissues. Consistent with this role, we and others have recently shown that YAP/TAZ promote vigorous proliferation of immature Schwann cells (SC) in developing peripheral nerves (Poitelon et al., 2016; Deng et al., 2017; Grove et al., 2017), and that overexpression of YAP/TAZ promotes abnormally excessive proliferation of mature SCs in adult peripheral nerves (Mindos et al., 2017; Wu et al., 2018). Unexpectedly, several groups demonstrated that YAP or YAP/TAZ promote differentiation of developing SCs by upregulating myelin-associated genes, thereby mediating developmental myelination (Fernando et al., 2016; Lopez-Anido et al., 2016; Poitelon et al., 2016; Deng et al., 2017; Grove et al., 2017). Our group additionally showed that YAP/TAZ are selectively expressed in differentiated myelin-forming SCs, and that they are required for maintenance of the myelin sheath in adult nerves (Grove et al., 2017). In many systems, YAP/TAZ shift to the cytoplasm concomitant with differentiation of developing cells, and the nuclear exclusion of YAP/TAZ is believed to be required for homeostatic maintenance of mature cells or tissues (Varelas, 2014; Wang et al., 2018). We were therefore intrigued to find that YAP/TAZ are nuclear and transcriptionally active in mature SCs maintaining peripheral myelination.

Building on these findings in developing and intact adult nerves, we now report on the role of YAP/TAZ in the regenerating nerve, in which SCs both proliferate and differentiate, as in developing peripheral nerve. Following traumatic nerve injury, SCs in axotomized nerve rapidly dedifferentiate and proliferate as they convert to regeneration promoting ‘repair’ SCs (Jessen and Mirsky, 2016; Tricaud and Park, 2017). When repair SCs regain axon contacts, they re-differentiate to myelin-forming SCs, thereby restoring motor and sensory functions (Fex Svennigsen and Dahlin, 2013; Stassart et al., 2013). Strikingly, we found that YAP/TAZ disappear from denervated SCs but reappear in SCs as axons regenerate. Consistent with these observations, we found that YAP/TAZ are dispensable for SC proliferation after injury but required for remyelination of regenerated axons. These findings extend the role of YAP/TAZ to functional regeneration of injured nerves and suggest that SCs are dependent on axons for their transcriptional activity of YAP/TAZ.

Results

YAP/TAZ expression in Schwann cells is axon-dependent

Transcriptional regulation of SC proliferation and differentiation by YAP/TAZ depends on their nuclear localization. Nuclear YAP/TAZ in SCs of adult mice promote myelin gene expression, essentially maintaining peripheral nerve myelination (Grove et al., 2017). As the first step to determine the roles of YAP/TAZ in nerve repair, we examined spatiotemporal expression patterns of YAP/TAZ in adult mice after sciatic nerve crush injury (Figure 1). The nerve crush model evokes active axon degeneration in the distal nerve stump, while permitting new axons from the proximal nerve stump to regenerate through the crushed site within 1–2 days post injury (dpi) (Kang and Lichtman, 2013; Jang et al., 2016; Frendo et al., 2019). New axons then keep regenerating within the basal lamina tubes filled with SCs and their processes, at the speed of 1–4 mm/day, although the debris of degenerating axons and myelin is not yet completely removed. We killed these mice 1, 3, 6, 9, 12, and 24 dpi and immunostained proximal and distal nerve stumps of ~5 mm in length with an antibody specific for both YAP and TAZ (Figure 1A; at 12 dpi). At one dpi when distal axons remained largely intact, nuclear expression of YAP/TAZ in associated SCs was unchanged (Figure 1B; 1D-dstl). Strikingly, at three dpi when axon degeneration was robust and SCs lost axon contacts, YAP/TAZ were almost undetectable in SC nuclei (Figure 1B; 3D-dstl). Thus, SCs lose nuclear expression of YAP/TAZ as associated axons degenerate.

Figure 1 with 1 supplement see all
Loss and recovery of YAP/TAZ in Schwann cells after sciatic nerve crush.

YAP/TAZ expression in crushed sciatic nerves of adult mice, shown by IHC (A, B) and Western blotting (C). Axons and Schwann cell (SC) nuclei are marked by neurofilament (NF) or Sox10, respectively. (A) A surgery schematic for nerve crush, which permits regeneration of axons into the distal nerve stump, illustrated by a low-magnification, longitudinal section of a sciatic nerve at 12 dpi, immunostained for YAP and TAZ. (B) Dramatic loss of YAP/TAZ in SC nuclei by three dpi, concomitant with axon degeneration, followed by upregulation of YAP/TAZ after six dpi, concomitant with axon regeneration. Right-most panels: zoomed area of merged images, as indicated, showing nuclear expression of YAP/TAZ in SCs associated with large diameter axons, before and after injury. Arrows point to large diameter axons in distal nerves before injury or 1 dpi, associated with YAP/TAZ+ SC nuclei. Arrowheads point to regenerating axons. Note that SC nuclei associated with a thin regenerating axon at 3 dpi do not express nuclear YAP/TAZ, but those in contact with a large diameter axon after 6 dpi do. Scale bars; 500 μm (A), 20 μm (B). (C) Western blotting of intact and crushed nerve lysates, showing loss of YAP and TAZ by 3 dpi, followed by full recovery of TAZ but not YAP by 12 dpi. Quantification of Western blots: n = 3–5 mice per experiment. ns = not significant, 2-way ANOVA. YAP: 1 dpi intact vs 1dpi crushed, p=0.9991; 1 dpi crushed vs 3 dpi crushed, ****p<0.0001; 1 dpi crushed vs 6 dpi crushed, ***p=0.0009; 1 dpi crushed vs 12 dpi crushed, ****p<0.0001; 3 dpi intact vs 3 dpi crushed, ****p<0.0001; 3 dpi crushed vs 6 dpi crushed, p=0.0652; 3 dpi crushed vs 12 dpi crushed, p=0.3479; 6 dpi intact vs 6 dpi crushed, ***p=0.0009; 6 dpi crushed vs 12 dpi crushed, **p=0.0018; 12 dpi intact vs 12 dpi crushed, ****p<0.0001. TAZ: 1 dpi intact vs 1 dpi crushed, p=0.9909; 1 dpi crushed vs 3 dpi crushed, ****P<0.0001; 1 dpi crushed vs 6 dpi crushed, p=0.6855; 1 dpi crushed vs 12 dpi crushed, p=0.9692; 3 dpi intact vs 3 dpi crushed, ****p<0.0001; 3 dpi crushed vs 6 dpi crushed, ****p<0.0001; 3 dpi crushed vs 12 dpi crushed, ****p<0.0001; 6 dpi intact vs 6 dpi crushed, p=0.9828; 6 dpi crushed vs 12 dpi crushed, p=0.9810; 12 dpi intact vs 12 dpi crushed, p>0.9999.

YAP/TAZ reappeared in the nuclei of SCs at 6 dpi, and these SCs were associated with regenerating axons (Figure 1B; 6D-dstl). By 12 dpi, as axon regeneration and maturation progressed further, more SCs exhibited strong nuclear expression of YAP/TAZ, comparable to that of SCs in the proximal neve stumps (Figure 1B; 12D-dstl, see also Figure 1A). These observations suggest that SCs upregulate nuclear YAP/TAZ, when they regain axon contacts as axons regenerate.

Notably, we frequently observed thin regenerating axons associated with SCs at 3dpi, but YAP/TAZ were undetectable in these SCs (Figure 1B; zoomed area of 3D-dstl). In contrast, SCs exhibiting strong YAP/TAZ at 6- and 12 dpi were associated with thick axons, which appeared large enough to be myelinated (i.e., 1>μm; Figure 1B; 6D-dstl, zoomed area of 12D-dstl). These observations suggest that YAP/TAZ are selectively upregulated in SCs associated with regenerating axons that have become large enough to be myelinated. Consistent with this notion, YAP/TAZ are expressed in myelinating, but not in non-myelinating, SCs (Grove et al., 2017).

Western blotting also revealed marked reduction of YAP and TAZ levels at 3 dpi, followed by rapid upregulation of TAZ levels (Figure 1C, Figure 1—source data 2). Notably, YAP levels remained low in nerve lysates at 12 dpi (Figure 1C). As cells other than SCs can affect overall YAP levels (Gaudet et al., 2011; Stierli et al., 2018), we next examined expression of YAP in SCs of crushed nerves by immunohistochemistry (IHC). We first verified that an antibody specifically recognized YAP, but not TAZ (Figure 1—figure supplement 1A). YAP is upregulated in many SC nuclei at 6 dpi and continues to be observed at 24 dpi (Figure 1—figure supplement 1B), demonstrating that YAP is also upregulated in SCs concomitant with axon regeneration.

The dramatic down- and upregulation of YAP/TAZ concurrent with axon degeneration and regeneration, respectively, suggest that SCs are dependent on axons for YAP/TAZ nuclear expression. To test further whether axons regulate YAP/TAZ expression in mature SCs, we next investigated if denervated SCs are capable of upregulating YAP/TAZ in the absence of regenerating axons. We used a nerve transection injury model in which we completely cut one sciatic nerve and tied both ends of the transected nerve to prevent axon regeneration from proximal to distal nerve stumps. We examined these mice at 1, 3, 6, 9, 12, and 24 dpi. We first confirmed absence of regenerating axons in distal nerve stumps of these mice and found that YAP/TAZ become undetectable in SCs at 3 dpi (Figure 2B; 3D-dstl), concurrent with robust axon degeneration as observed after nerve crush injury. Notably, nuclear YAP/TAZ continued to be undetectable in SCs after 3 dpi (Figure 2B; e.g., 12D-dstl, see also Figure 2A), demonstrating that axons are required for YAP/TAZ upregulation in denervated SCs.

YAP/TAZ expression in Schwann cells after sciatic nerve transection.

(A, B, C) YAP/TAZ expression in transected sciatic nerves of adult mice. Axons and Schwann cell (SC) nuclei are marked by neurofilament (NF) or Sox10, respectively. (A) A surgery schematic for nerve transection illustrated by a low-magnification, longitudinal section of a sciatic nerve at 12 dpi, immunostained for YAP and TAZ. Axon regeneration into the distal nerve stump was prevented by ligating the transected nerve stumps. (B) Complete loss of YAP/TAZ in SC nuclei at and after 3 dpi, concomitant with axon degeneration. Right-most panels: zoomed area of merged images, as indicated, showing that SCs do not upregulate YAP/TAZ in the absence of regenerating axons. (C) Cytoplasmic loss of phosphorylated YAP (p-YAP) in SCs of transected nerve. p-YAP was undetectable in axotomized SCs at 12 dpi. (D) Upregulation of p-YAP in SCs of crushed nerve. p-YAP was detectable in innervated SCs at 12 dpi. Right-most panel: zoomed area of merged image, showing a SC nucleus exhibiting perinuclear cytoplasmic p-YAP. Scale bars; 500 μm (A), 20 μm (B–D).

We also used an antibody specific for transcriptionally inactive, phosphorylated YAP (p-YAP), which is located preferentially in cytoplasm and exhibits perinuclear and membrane accumulation (Grove et al., 2017). We found that p-YAP became undetectable in SCs of transected/tied nerves by 12 dpi (Figure 2C). In contrast, p-YAP expression recovered in SCs of crushed nerves at 12 dpi (Figure 2D). These findings suggest that SCs are dependent on axons for both nuclear and cytoplasmic expression of YAP/TAZ.

YAT/TAZ are dispensable for Schwann cell proliferation after nerve injury

SCs rapidly dedifferentiate and convert to repair SCs after nerve injury. During this transdifferentiation process, SCs begin to proliferate ~3 dpi (Clemence et al., 1989; Jessen and Mirsky, 2016; Tricaud and Park, 2017). Our observation that YAP/TAZ disappear in SCs by 3 days after axotomy raises the interesting possibility that YAP/TAZ are not involved in injury-elicited SC proliferation. Alternatively, levels of YAP/TAZ that are too low to be detected by IHC may be sufficient to promote transcription of the genes activating SC proliferation. To test these possibilities, we used an inducible knockout mouse (Plp1-creERT2; Yapfl/fl; Tazfl/fl, hereafter Yap/Taz iDKO) to inactivate YAP/TAZ selectively in SCs after nerve injury. We induced recombination at 6 weeks of age, completely transected and tied the sciatic nerve in one leg, killed the mice 5 days later when SCs actively proliferate, and compared SC proliferation in intact and transected nerves of WT and iDKO mice (Figure 3A, n = 3 mice per genotype). We first confirmed efficient ablation of YAP/TAZ in SCs by analyzing contralateral, intact nerves of iDKO mice (Figure 3B and F). We excluded mice with poor deletion (i.e., exhibiting YAP/TAZ in >20% SCs) from further analysis. Notably, pulse labeling with EdU indicated that the transected nerves of WT and iDKO contained similar numbers of dividing SCs in S phase (Figure 3C and G). Numbers of Ki67+ proliferating SCs (Figure 3D and H) and of total SCs (Figure 3E and I) were also similar in the transected nerves of WT and iDKO.

Figure 3 with 1 supplement see all
YAP/TAZ are dispensable for Schwann cell proliferation after axotomy.

(A) Schematic showing timeline of tamoxifen injection, sciatic nerve transection and sacrifice of adult WT or Yap/Taz iDKO. (B) Longitudinal sections of intact sciatic nerves showing efficient deletion of YAP/TAZ in iDKO. SC nuclei are marked by Sox10 (red). All cell nuclei are marked by DAPI (blue). (C) Longitudinal sections of transected nerves of WT or iDKO showing SCs in S-phase of the cell cycle marked by EdU (green). (D) Longitudinal sections of transected nerves of WT or iDKO showing proliferating SCs marked by Ki67 (green). (E) Transverse sections of transected nerves of WT or iDKO showing SCs marked by Sox10 (red). (F) Quantification of SCs expressing nuclear YAP/TAZ in intact sciatic nerves of WT or iDKO. n = 3 mice per genotype, *p=0.0495, Mann-Whitney. (G) Quantification of EdU+ SCs in transected nerves of WT or iDKO. n = 3 mice per genotype, *p=0.0463, Mann-Whitney. (H) Quantification of Ki67+ proliferating SCs in transected nerves of WT or iDKO. n = 3 mice per genotype, ns, not significant, p=0.5127, Mann-Whitney. (I) Quantification of Sox10+ SCs in transected nerves of WT or iDKO. n = 3 mice per genotype. ns, not significant, p=0.8273, Mann-Whitney. Scale bars = 30 μm (B–E).

If adult SCs lacking YAP/TAZ in iDKO die or proliferate independently of axotomy, our analysis of injury-elicited SC proliferation might be confounded. To exclude this possibility, we examined contralateral, intact nerves of WT and iDKO mice at 12 dpi for SC proliferation and death (Figure 3—figure supplement 1A). Contralateral iDKO nerves contained neither EdU+ SCs (Figure 3—figure supplement 1B) nor apoptotic SCs, as assessed by TUNEL assays (Figure 3—figure supplement 1C). We also found that SC numbers did not differ significantly from those in intact nerves of WT mice (Figure 3—figure supplement 1D and E). Collectively, these results strongly indicate that YAP/TAZ do not regulate SC proliferation after nerve injury.

SCs lacking YAP/TAZ convert to repair SCs and support axon regeneration

Next, we investigated if transdifferentiation to repair SCs proceeds normally in iDKO nerves after injury. We first examined expression of c-Jun, phosphorylated c-Jun (pc-Jun), p75 and Oct-6, which are associated with formation of repair SCs during nerve regeneration (Scherer et al., 1994; Parkinson et al., 2008; Arthur-Farraj et al., 2012; Fontana et al., 2012). Repair SC formation principally depends on the upregulation of c-Jun, which promotes expression of regeneration-associated genes (RAG), such as p75 neurotrophin receptor (NTR) (Parkinson et al., 2008; Arthur-Farraj et al., 2012; Fontana et al., 2012). Immunohistochemical analysis of transected sciatic nerves at five dpi showed that cJun, pc-Jun, p75 NTR and Oct-6 were all upregulated in denervated SCs of iDKO mice after nerve injury, as in WT mice (Figure 4, Figure 4—figure supplement 1 and Figure 4—source data 3). Notably, injured nerves of WT and iDKO mice contained similar numbers of SCs expressing c-Jun (Figure 4A and E) and active pc-Jun (Figure 4B and F). There was minimal expression of c-Jun in contralateral, intact nerves of iDKO at five dpi (Figure 4—figure supplement 1). p75 NTR expression was also strongly upregulated in iDKO SCs, as in WT SCs (Figure 4C and G), and Oct-6 expression in WT and iDKO SCs did not differ (Figure 4D and H). Western blotting analysis confirmed upregulation of these proteins in injured nerves and validated the specificity of the antibodies used in the immunohistochemical analysis (Figure 4—figure supplement 1 and Figure 4—source data 3).

Figure 4 with 1 supplement see all
Schwann cells lacking YAP/TAZ transdifferentiate into repair Schwann cells.

Longitudinal sections of transected sciatic nerves of WT and Yap/Taz iDKO immunostained by various markers of growth-promoting repair SCs at five dpi. SCs are marked by Sox10 (red). (A) Representative sections showing upregulation of c-Jun in iDKO SCs, as in WT SCs. (B) Upregulation of active phospho-S63 c-Jun in iDKO SCs, as in WT. (C) Upregulation of p75 in iDKO SCs, as in WT SCs. (D) Upregulation of Oct-6 in iDKO SCs, as in WT SCs. (E) Quantification of c-Jun+ SCs in WT and iDKO. n = 3 mice per genotype. ns, not significant, p=0.1266, Mann-Whitney. (F) Quantification of pc-Jun+ SCs in WT and iDKO. n = 3 mice per genotype. *p=0.0495, Mann-Whitney. (G) Quantification of p75+ SCs in WT and iDKO. n = 3 mice per genotype. ns, not significant, p=0.5127, Mann-Whitney. (H) Quantification of Oct-6+ SCs in WT and iDKO. n = 3 mice per genotype. ns, not significant, p=0.8273, Mann-Whitney. Scale bars = 30 μm (A–D).

SCs are essential for successful nerve regeneration (Scheib and Höke, 2013; Jessen and Mirsky, 2016). As the definitive test of whether iDKO SCs convert normally to repair SCs, we next examined if the absence of YAP/TAZ in SCs impairs nerve regeneration. Because Yap/Taz iDKO mice die ~14 days after tamoxifen treatment (Grove et al., 2017), we crushed sciatic nerves and analyzed them on 12–13 dpi. To minimize variability, we crushed nerves at the same site close to the sciatic notch and analyzed nerve segments immunohistochemically or ultrastructurally at the same distance distal to the injury (Figure 5A). An anti-β3 tubulin antibody, which identifies all axons, intensely labeled many axons that had regenerated through the ~10 mm long distal nerve stumps of iDKO mice (Figure 5B and D). These axons were as thick and numerous in iDKO as in WT nerves (Figure 5B and F, see also Figure 7—figure supplement 1). Similar numbers of axons were also present in contralateral intact nerves of WT and iDKO (Figure 7—figure supplement 1), indicating that there was no axon degeneration in intact nerves of iDKO at 12–13 dpi.

Schwann cells lacking YAP/TAZ support axon regeneration.

(A) Schematic showing relative locations and sizes of the distal nerve segments used for ultrastructural or light microscopic analysis of axon regeneration in WT or Yap/Taz iDKO, 12–13 days after nerve crush. (B) Low magnification views of longitudinal sections of ~5 mm long nerve segments distal to the crush site, showing regenerated axons in iDKO as abundant as in WT. Axons are marked by TuJ1. (C, D) High magnification views of boxed area in (B), ~8 mm distal to the crush site. (E) Low and high magnification views of TEM, taken at 5 mm distal to the crush site, showing numerous axons that regenerated within basal lamina tubes in iDKO, as in WT. ‘ax’ denotes an axon. Numerous axons are large (>1 μm) but unmyelinated in iDKO. Examples of single large myelinated axons in WT (E–a, E–b), single large unmyelinated axon in iDKO (E–c) and axon bundles containing a large unmyelinated axon in iDKO (E–d). (F) Quantification of the axon density in crushed nerves of WT and iDKO, n = 4 mice for WT and three mice for iDKO. ns, not significant, p=0.4715, Mann-Whitney. (G) Quantification of the percentage of BL tubes containing axons in crushed nerves of WT and iDKO, n = 4 mice for WT and three mice for iDKO. ns, not significant, p=0.7237, Mann-Whitney (H) Quantification of the percentage of BL tubes containing at least one axon >1 μm in diameter, in crushed nerves of WT and iDKO. n = 4 mice for WT and three mice for iDKO. ns, not significant, p=0.1573, Mann-Whitney. (I) Quantification of the percentage of BL tubes containing multiple axons, at least one of which is >1 μm in diameter, in crushed nerves of WT and iDKO. n = 4 mice for WT and three mice for iDKO. *p=0.0339, Mann-Whitney. Scale bars = 500 μm (B), 100 μm (C, D), 2 μm (E).

To confirm these findings, we examined transverse nerve segments 5 mm distal to the injury by TEM (Transmission Electron Microscopy). In this ultrastructural analysis, we took advantage of the fact that regenerating axons extend through the basal lamina (BL) tubes that surround SCs and their processes (Scheib and Höke, 2013; Jessen and Mirsky, 2016). We found that the percentage of BL tubes containing axons (single or multiple) was similar in WT and iDKO nerves (Figure 5E and G). Furthermore, the percentage of BL tubes containing axons large enough to be myelinated (i.e.,>1 μm) did not differ (Figure 5H). However, the large axons in iDKO nerves were more frequently accompanied by one or multiple, often thin, axons, which presumably represent transient collateral sprouts (Figure 5E–d and I).

Next, we examined iDKO nerves at an earlier time point after injury to investigate if axon regeneration might be delayed in the absence of YAP/TAZ in SCs. We analyzed longitudinal sections of crushed sciatic nerves from WT and iDKO mice at three dpi (Figure 6; n = 3 mice per genotype). Because abundant debris of degenerating axons often confounded immunohistochemical identification of regenerating axons at this early time point, we selectively labeled regenerating axons with an antibody for superior cervical ganglion 10 (SCG10), which is rapidly and preferentially upregulated in sensory axons after injury (Shin et al., 2014; Mogha et al., 2016). Numerous axons reached ~4 mm distal to the crush site in iDKO, as in WT mice (Figure 6A,B). There was no significant difference in axon density measured 2 mm distal to the crush (Figure 6C–E), nor in the length of the longest axons (Figure 6F). Taken together, these results show that SCs lacking YAP/TAZ convert normally to repair SCs and support axon regeneration after injury.

Schwann cells lacking YAP/TAZ support timely axon regeneration after acute injury.

(A) Schematic showing relative locations of crushed site, axon quantification and sizes of the distal nerve segments used for light microscopic analysis of axon regeneration in WT or Yap/Taz iDKO, 3 days after nerve crush. (B) Low magnification views of longitudinal sections, showing abundant axon regeneration in both WT and iDKO. Regenerating axons are marked by SCG10. (C, D) High magnification views of boxed areas in (B), showing numerous thin regenerating axons. (E) Quantification of the axon density measured at 2 mm distal to the crushed site. n = 3 mice per genotype. ns, not significant, p=0.2752, Mann-Whitney. (F) Quantification of the distance regenerated by the longest axon. n = 3 mice per genotype. ns, not significant, p=0.8273, Mann-Whitney. Scale bars = 1 mm (B), 100 μm (C, D).

YAP/TAZ are required for Schwann cells to remyelinate axons

We have previously reported that developing SCs lacking YAP/TAZ arrest as promyelinating SCs, and are therefore unable to initiate myelin formation (Grove et al., 2017). To determine if adult SCs lacking YAP/TAZ can myelinate regenerating axons, we next analyzed the extent of myelination in the same iDKO nerves analyzed for axon regeneration on 12–13 dpi (Figure 7A shows the same nerves as Figure 5B). As expected, there was strong expression of myelin basic protein (MBP), a major structural component of the myelin sheath, in the crushed nerves of WT mice (Figure 7A and B). MBP immunoreactivity was also abundant in the contralateral, intact nerves of iDKO mice (Figure 7A; bottom panel), in which our previous ultrastructural analysis found segmental demyelination (Grove et al., 2017). In contrast, iDKO crushed nerves revealed remarkably little, if any, MBP immunoreactivity (Figure 7A and C, see also Figure 7—figure supplement 1 for higher magnification images). Consistent with this immunohistochemical analysis, semithin (Figure 7D) and ultrathin sections processed for EM (Figures 5E and 7E) contained many myelinated axons in WT but almost none in iDKO crushed nerves (Figure 7F and G). Moreover, iDKO SCs frequently surrounded and established a 1:1 relationship with large axons, but none of these axons exhibited a myelin sheath (Figures 5E and 7E). These findings suggest that adult SCs lacking YAP/TAZ fail to remyelinate axons because they arrest at the promyelinating stage after injury.

Figure 7 with 1 supplement see all
Schwann cells lacking YAP/TAZ fail to myelinate regenerated axons.

Ultrastructural and light microscopic analyses of remyelination in distal nerves of WT or Yap/Taz iDKO, 12–13 days after nerve crush. (A) Low magnification views of longitudinal sections of intact or crushed nerves of WT and iDKO, showing no myelination of regenerated axons in crushed nerves of iDKO as indicated by the lack of MBP immunostaining. Refer to Figure 5B for robustly regenerated axons in the same iDKO mouse. (B, C) High magnification views of boxed area in (A), showing abundant regenerated axons in crushed nerves of both WT (B) and iDKO (C). Note that regenerated axons in iDKO are not myelinated. Axons and myelin are marked by TuJ1 and MBP, respectively. (D) Semi-thin sections stained with toluidine blue showing numerous myelinated axons in crushed nerves of WT but not in iDKO. (E) TEM images of representative single large axons, myelinated in WT (left panel) but unmyelinated in iDKO (right panel). (F) Quantification of the percentage of single axons that are myelinated. n = 4 mice for WT and three mice for iDKO. *p=0.0323, Mann-Whitney. (G) G-ratio in WT and iDKO. Myelinated axons in WT are compared to unmyelinated single axons in iDKO. n = 3 mice per genotype. *p=0.0495 Mann-Whitney. Scale bars = 500 μm (A), 100 μm (B, C), 10 μm (D), 2 μm (E).

YAP and TAZ are functionally redundant and required for optimal remyelination

Mindos et al., 2017 recently reported that expression of YAP, assessed by Western blotting, selectively increases after nerve injury in mutant nerves lacking Merlin in SCs, but not in WT nerves, whereas TAZ increases in both WT and mutant nerves. They also reported that elevated YAP levels prevent axon regeneration and remyelination, and that inactivation of YAP alone is sufficient to restore full functional recovery of the Merlin mutants (Mindos et al., 2017). These observations suggest that the function of TAZ in adult SCs may differ from that of YAP. We next examined axon regeneration and remyelination when SCs express YAP but not TAZ after injury. We reasoned that, if YAP prevents regeneration, regardless of expression levels (see Discussion), and if it differs functionally from TAZ, then we would find axon regeneration and remyelination to be poor.

Using a TAZ-selective tamoxifen inducible line to inactivate TAZ in SCs (Plp1-creERT2; Yap+/+; Tazfl/fl, hereafter Taz iKO), we crushed sciatic nerves unilaterally and compared the mutants to WT and Yap/Taz iDKO mice at 12 dpi. We first confirmed efficient ablation of TAZ and no compensatory elevation of YAP levels in Taz iKO (Figure 8A, Figure 8—source data 4). We then used anti-β3 tubulin antibody to assess axon regeneration up to 15 mm distal to the crushed site (Figure 8—figure supplement 1A,B). We found that regenerating axons were as thick and numerous in Taz iKO, as in WT mice (Figure 8—figure supplement 1C,D). Axon density measured at 8 ~ 10 mm distal to the crush site showed no significant difference among WT, Taz iKO and Yap/Taz iDKO nerves (Figure 8B, Figure 8—figure supplement 1E).

Figure 8 with 1 supplement see all
YAP and TAZ are redundantly required for optimal remyelination.

(A) Western blotting of intact sciatic nerve lysates, showing markedly reduced TAZ in Taz iKO, whereas YAP levels remain relatively unchanged. YAP band is tighter and faster migrating in Taz iKO, than in WT, indicative of reduced phosphorylation. Quantification of Yap and Taz in WT and Taz iKO, n = 3 mice per genotype. YAP: ns, not significant, p=0.2752, Mann-Whitney. TAZ: *p=0.0495, Mann-Whitney. (B) Quantification of axon density in WT, Yap/Taz iDKO and Taz iKO nerves at 12 dpi, 8–10 mm distal to crush site (also see Figure 5B,F and Figure 8—figure supplement 1B, (E). n = 4 mice for WT, three mice for iDKO and Taz iKO: WT vs iDKO, p=0.72; WT vs iKO, p=0.41; iDKO vs iKO, p=0.18, all not significant, one-way ANOVA with Tukey’s multiple comparison test. (C–G) Comparative analysis of axon regeneration and remyelination in WT and Taz iKO, 12–13 days after nerve crush. (C) Representative TEM images of WT and Taz iKO nerves, taken at 5 mm distal to the crush site, showing numerous axons that regenerated within basal lamina tubes in Taz iKO, as in WT. ‘ax’ denotes a single axon. Some large axons are myelinated in Taz iKO. (D) Quantification of the percentage of BL tubes containing axons of any diameter in WT, Taz iKO and Yap/Taz iDKO nerves. n = 4 mice for WT, three mice for iDKO and two mice for Taz iKO: WT vs. iDKO, p=0.99; WT vs. iKO, p=0.90; iDKO vs. Taz iKO, p=0.92, all not significant, one-way ANOVA with Tukey’s multiple comparison test. (E) Quantification of the percentage of BL tubes containing at least one axon larger than 1 μm in diameter in WT, Taz iKO and Yap/Taz iDKO nerves. n = 4 mice for WT, three mice for iDKO and two mice for Taz iKO: WT vs. iDKO, p=0.73; WT vs. iKO, p=0.22; iDKO vs. iKO, p=0.52, all not significant, one-way ANOVA with Tukey’s multiple comparison test. (F) Quantification of the percentage of single axons that are remyelinated in WT, Taz iKO and Yap/Taz iDKO nerves. n = 4 mice for WT, three mice for iDKO and two mice for Taz iKO: WT vs. iDKO, ****p<0.0001; WT vs. iKO, **p=0.0094; iDKO vs. Taz iKO, **p=0.0016, one-way ANOVA with Tukey’s multiple comparison test. (G) G-ratios of remyelinated axons in WT and Taz iKO nerves, compared to unmyelinated axons in Yap/Taz iDKO nerve. WT and Taz iKO remyelinated axons have equivalent G-ratios. n = 6 mice for WT, three mice for iDKO and two mice for iKO: WT vs. iDKO, ****p<0.0001; WT vs. iKO, not significant, p=0.074; iDKO vs. iKO, ***p=0.0008, one-way ANOVA with Tukey’s multiple comparison test. Scale bar = 2 μm (C).

Ultrastructural analysis of nerve segments at 5 mm distal to the injury revealed many BL tubes containing single or multiple axons in Taz iKO, as in WT (Figure 8C). These axon-containing BL tubes were as numerous in iKO as in WT and iDKO (Figure 8D). Counts of BL tubes containing axons large enough to be myelinated also did not differ (Figure 8E). Taken together, these results show that axons regenerated as robustly in Taz iKO as in WT and iDKO nerves, indicating that SCs expressing only YAP supported axon regeneration.

We also found that, whereas iDKO nerves contained no myelinated axons (e.g., Figure 7D), myelinated axons were frequent in Taz iKO nerves (Figure 8C and F), and G-ratios did not differ in Taz iKO and WT (Figure 8G), demonstrating that SCs expressing only YAP were capable of myelinating regenerated axons. Notably, however, a significantly smaller percentage of single axons were myelinated in Taz iKO than in WT (Figure 8F), indicating that remyelination is less advanced in Taz iKO nerves whose SCs express only YAP. Taken together, these results show that YAP, at normal levels (see Discussion), does not prevent axon regeneration or remyelination after injury, and that both YAP and TAZ are required for optimal remyelination.

Redifferentiation of Schwann cells lacking YAP/TAZ

Following axon regeneration, denervated SCs that have regained axon contacts downregulate dedifferentiation-associated genes while upregulating genes promoting their differentiation (Stassart et al., 2013; Quintes et al., 2016; Wu et al., 2016). It is possible that YAP/TAZ-deficient iDKO SCs fail to myelinate regenerated axons because their capacity to carry out one or both processes is defective. To test if iDKO SCs correctly downregulate dedifferentiation-associated genes, we compared expression of c-Jun, Ki67 and Oct-6 by WT and iDKO SCs at 5 and 12 dpi after crush. The number of c-Jun+ SCs was markedly, but similarly, reduced in nerves of both WT and iDKO at 12 dpi (Figure 9A,E), and proliferating SCs were rare (Figure 9B,F). Oct-6 expression was also reduced in both WT and iDKO (Figure 9C,G), although it remained statistically higher in iDKO SCs. These results suggest that iDKO SCs are capable of downregulating dedifferentiation genes and withdraw gradually from dedifferentiation as like WT SCs.

Redifferentiation of Schwann cells lacking YAP/TAZ.

Longitudinal sections of crushed nerves of WT and Yap/Taz iDKO at 12 dpi, immunostained by various markers of SC dedifferentiation (c-Jun and Oct-6), proliferation (Ki67) and redifferentiation (Krox20). SCs are marked by Sox10. (A) Representative sections showing c-Jun+ SCs markedly reduced in iDKO, as in WT. (B) Representative sections showing rarely observed Ki67+ proliferating SCs in iDKO, as in WT. (C) Representative sections showing Oct-6+ SCs reduced in iDKO, as in WT. (D) Representative sections showing failed upregulation of Krox20 in iDKO SCs. (E) Quantitative comparison of c-Jun+ SCs at 5 and 12 dpi, showing similar downregulation of c-Jun in WT and iDKO SC. n = 3 mice per genotype, 2-way ANOVA, ns = not significant. WT five dpi vs WT 12 dpi, **p=0.0069; WT five dpi vs iDKO five dpi, p=0.4260; WT 12 dpi vs iDKO 12 dpi, p=0.9574; iDKO five dpi vs iDKO 12 dpi, **p=0.0018. (F) Quantitative comparison of Ki67+ SCs, showing similar reduction in proliferating SCs in WT and iDKO nerves between 5 dpi and 12 dpi. n = 3 mice per genotype, 2-way ANOVA, ns = not significant. WT five dpi vs WT 12 dpi, ****p<0.0001; WT five dpi vs iDKO five dpi, p>0.9999; WT 12 dpi vs iDKO 12 dpi, p=0.6775; iDKO five dpi vs iDKO 12 dpi, ****p<0.0001. (G) Quantitative comparison of Oct-6+ SCs, showing significant downregulation of Oct-6 in WT and iDKO SCs between 5 dpi and 12 dpi. n = 3 mice per genotype, ns = not significant, 2-way ANOVA. WT five dpi vs WT 12 dpi, ***p=0.0005; WT five dpi vs iDKO five dpi, p=0.9817; WT 12 dpi vs iDKO 12 dpi, *p=0.0221; iDKO five dpi vs iDKO 12 dpi, *p=0.0299. (H) Quantitative comparison of Krox20+ SCs, showing upregulation of Krox20 in WT SCs, but not in iDKO SCs between 5 dpi and 12 dpi. n = 3 mice per genotype, 2-way ANOVA, ns = not significant. WT five dpi vs WT 12 dpi, ****p<0.0001; WT five dpi vs iDKO five dpi, p>0.9999; WT 12 dpi vs iDKO 12 dpi, ****p<0.0001; iDKO five dpi vs iDKO 12 dpi, p>0.9999. Scale bar = 10 μm (A–D).

Lastly, we examined expression of Krox 20 (also known as Egr2), the master transcription factor that drives myelin gene expression (Topilko et al., 1994; Decker et al., 2006). Notably, whereas WT SCs upregulated Krox 20 expression at 12 dpi, concomitant with remyelination, few if any iDKO SCs exhibited Krox 20 immunoreactivity (Figure 9D,H). These results suggest that iDKO SCs fail to myelinate regenerated axons at least in part due to failure to upregulate Krox 20.

Discussion

Recent studies of SC-specific gene targeting consistently show that SCs lacking both YAP and TAZ are unable to proliferate properly and fail to myelinate developing peripheral nerves (Poitelon et al., 2016; Deng et al., 2017; Grove et al., 2017). It remains controversial, however, how YAP/TAZ loss results in complete amyelination of developing nerves and whether YAP/TAZ also play a role in myelin maintenance of adult nerves. Indeed, Poitelon et al., attributed developmental amyelination to the inability of immature SCs lacking YAP/TAZ to wrap around developing axons, a process called radial sorting (Feltri et al., 2016; Poitelon et al., 2016). In contrast, Grove et al. and Deng et al. attributed the myelination failure primarily to the inability of SCs to differentiate into myelinating SCs (Deng et al., 2017; Grove et al., 2017). Deng et al., however, disagreed with our view about the role of YAP/TAZ in myelin maintenance of adult nerves (Deng et al., 2017; Grove et al., 2017). These disagreements motivated the present study, which investigated YAP/TAZ expression in adult SCs after nerve injury and their contribution to nerve regeneration. We found that YAP/TAZ dramatically disappear and reappear in SCs after nerve injury, and this loss and recovery of YAP/TAZ in SCs are spatiotemporally correlated with degeneration and regeneration of axons. We also found that SCs lacking YAP/TAZ proliferate and wrap around regenerated axons normally, but then fail to remyelinate them. These findings have several important implications for YAP/TAZ function in mature SCs.

Using antibodies specifically immunolabeling YAP or YAP/TAZ, we found dramatic down- and upregulation of both nuclear and cytoplasmic YAP/TAZ in SCs after nerve injury. Immunohistochemical identification of SC-selective YAP/TAZ was essential for detecting spatiotemporal regulation of YAP/TAZ. Indeed, we were only able to detect YAP/TAZ downregulation on Western blots when we used lysates prepared from nerves extensively perfused with saline, and from which the epi- and perineurium had been carefully removed. This procedure probably succeeded because it minimized the amount of YAP/TAZ present in cells other than SCs. Careful attention to YAP/TAZ expression in cells other than SCs will help to resolve inconsistencies in earlier studies of YAP/TAZ expression in peripheral nerve.

YAP/TAZ are located in the nuclei of developing SCs, where they promote proliferation and differentiation (Poitelon et al., 2016; Deng et al., 2017; Grove et al., 2017). They are also nuclear in adult SCs that maintain the myelin sheath (Grove et al., 2017) and that proliferate abnormally (Wu et al., 2018). It was therefore particularly intriguing to find that YAP/TAZ become undetectable in denervated SCs and that SCs lacking YAP/TAZ proliferate normally. YAP/TAZ disappeared from SCs, upon axon degeneration in both crushed and transected nerves. They reappeared in SCs in crushed nerve concomitant with regenerating axons, but not in transected nerve lacking axons, suggesting that YAP/TAZ expression in SCs is dependent on axons. It is also notable that YAP/TAZ appeared unchanged at one dpi, but had dramatically disappeared at three dpi, when axon degeneration was well underway (Beirowski et al., 2005; Gomez-Sanchez et al., 2015; Jang et al., 2016) and denervated SCs had lost contact with axons. Furthermore, SCs that upregulated YAP/TAZ after three dpi were associated with regenerating axons large enough to be myelinated. These results are in consistent with our earlier findings of selective expression of YAP/TAZ in myelin-forming SCs (Grove et al., 2017), implying that direct SC-axon contact probably regulate YAP/TAZ down- and upregulation after nerve injury.

We were surprised to find that SC proliferation proceeds normally in Yap/Taz iDKO nerves after injury. Proliferation of mature SCs after injury is therefore due to a YAP/TAZ-independent mechanism, in contrast to the proliferation of developing SCs, which is markedly reduced by YAP/TAZ inactivation (Clemence et al., 1989; Grove et al., 2017). This result is consistent with the notion that the mechanism for SC proliferation during development differs from that for proliferation after injury (Atanasoski et al., 2001; Atanasoski et al., 2008). However, our finding does not indicate that YAP/TAZ are unable to stimulate proliferation of mature SCs. Abnormally high levels of YAP have been shown to elicit excessive SC proliferation in Merlin mutants after nerve injury (Mindos et al., 2017), and YAP/TAZ overexpression induced by LATS1/2 inactivation has been shown to induce tumorigenic SC proliferation in adult nerves (Wu et al., 2018). These observations, together with our own, indicate that YAP/TAZ are not normally involved in injury-elicited SC proliferation, but that, if abnormally overexpressed, they can stimulate vigorous SC proliferation. It is also noteworthy that YAP/TAZ inactivation markedly reduces, but does not completely prevent, proliferation of developing SCs (Deng et al., 2017; Grove et al., 2017). We suggest, therefore, that YAP/TAZ are potent stimulants of SC proliferation, but not an absolute requirement.

Tumorigenic proliferation of adult SCs associated with abnormally increased YAP/TAZ levels (Wu et al., 2018) suggests the importance of maintaining proper levels of YAP/TAZ, but it does not explain why YAP/TAZ are almost completely lost, rather than reduced, in denervated SCs. We previously demonstrated that inducible deletion of YAP/TAZ elicits SC demyelination in adult intact nerve (Grove et al., 2017). If YAP/TAZ indeed maintain myelination and act by promoting transcription of Krox20 and other myelin genes, then sustaining YAP/TAZ would counteract demyelination and dedifferentiation of SCs after injury. Conversely, their absence would promote downregulation of myelin genes, facilitating demyelination and formation of repair SCs. In accordance with these ideas, transcription of Krox 20 and other myelin genes remains robust in SCs up until two dpi, but is downregulated by three dpi (Arthur-Farraj et al., 2017), when we observed dramatic disappearance of YAP/TAZ. This timing suggests that YAP/TAZ protein downregulation leads to Krox 20 mRNA downregulation, suppressing expression of myelin proteins in de-differentiating Schwann cells. Complete loss of both nuclear and cytoplasmic YAP/TAZ could therefore imply that active regulatory mechanisms completely inactivate YAP/TAZ after injury. In support of this notion, in mutant mice lacking Merlin, YAP is abnormally upregulated in SCs after nerve injury, impairing SC de-differentiation (Mindos et al., 2017). This YAP upregulation suggests that one role of Merlin is to downregulate YAP in SCs and that YAP/TAZ expression in SCs is likely under active, presumably axon-dependent, regulation in both intact and injured mature nerves. This postulation of axon-dependent regulation of YAP/TAZ emphasizes that YAP/TAZ play a passive role in Wallerian degeneration, predicting that SCs do not require YAP/TAZ to dedifferentiate, proliferate, or transdifferentiate to repair SCs. Indeed, we found that these processes proceed normally in Yap/Taz iDKO mice. iDKO SCs are capable of wrapping around large diameter single axons but fail to initiate remyelination, which recapitulates the developmental phenotype of these mutant mice (Deng et al., 2017; Grove et al., 2017). For at least two reasons remyelination failure is highly unlikely to be due to poor physiological conditions of iDKO mice that die ~14 days after injury. First, axons regenerate normally in iDKO, which is unlikely if SCs are selectively vulnerable to poor physiological condition. Furthermore, iDKO SCs proliferate and trans-differentiate to repair-SCs normally. iDKO SCs also downregulate c-Jun to prepare for remyelination, whereas they maintain a higher level of Oct six than WT SCs, consistent with the failure of iDKO SCs to upregulate Krox 20 and MBP. Second, iDKO SCs wrap around individual axons, but fail to myelinate them, indicating that they proceed to the promyelination stage but no further. Therefore, one would have to postulate that the poor physiological condition of iDKO mice has a very specific effect on a particular remyelination stage, which we find unlikely.

iDKO SCs fail to upregulate Krox 20. Krox 20 is widely accepted as the key transcription factor promoting peripheral myelination. This is largely believed to be the case after injury (Brügger et al., 2017), although its role in remyelination has never been explicitly demonstrated. Other pathways can also promote upregulation of certain myelin proteins and lipids independently of Krox 20 (Domènech-Estévez et al., 2016), and numerous other factors, both positive and negative regulators, mediate peripheral myelination (Jessen and Mirsky, 2008; Herbert and Monk, 2017). Therefore, it is conceivable that the lack of Krox 20 in iDKO is a consequence rather than a cause of impaired remyelination. However, several considerations make it highly unlikely. First, we and others demonstrated that YAP/TAZ-TEAD1 complex directly binds to the cis-acting regulatory sequence of Krox 20 designated as the Myelinating Schwann cell Element (MSE) to upregulate Krox 20 during developmental myelination (Lopez-Anido et al., 2016; Grove et al., 2017). Second, we showed that YAP/TAZ-TEAD1 also bind to Krox 20 MSE in adult nerve, suggesting direct regulation of Krox 20 by YAP/TAZ in mature SCs (Grove et al., 2017). Third, Oct 6, which induces upregulation of Krox 20, together with other TFs, is upregulated in iDKO after nerve injury, as in WT (Figures 4H and 9G). This finding suggests that remyelination in iDKO is blocked at the step of Krox 20 upregulation. Indeed, we and others have shown that iDKO SCs are arrested at the promyelination stage during development (Deng et al., 2017; Grove et al., 2017), and we found that, similarly after nerve injury, iDKO SCs proceed to the promyelination stage but fail to upregulate Krox 20 and initiate myelin formation.

YAP upregulation in SCs lacking Merlin has recently been reported to decrease the regeneration-promoting ability of repair SCs, which prevents axon regeneration in Merlin mutants (Mindos et al., 2017). This study implicates YAP as an inhibitor of axon regeneration. Our study suggests that this inhibition is dose- and context-dependent. We observed that repair SCs rapidly upregulate both YAP and TAZ as axons regenerate and that expression persists as regeneration continues. We also found that axon regeneration is as robust in Taz iKO and Yap/Taz iDKO as in WT, but not noticeably enhanced. Given that YAP is not compensatorily upregulated in Taz iKO (Figure 8A), these results suggest that at least normal levels of YAP do not prevent axon regeneration. However, overly robust upregulation of YAP, presumably as in Merlin mutants (Mindos et al., 2017), may severely compromise axon regeneration because excessive levels of YAP/TAZ alter the growth-promoting ability of SCs and/or cause their tumorigenic proliferation.

The present study, together with earlier work, strongly suggests that the levels of YAP/TAZ may be a critical determinant of their function in adult SCs. Optimal levels of YAP/TAZ promote myelin formation, maintenance and remyelination, whereas their absence promotes demyelination. In contrast, overly excessive levels of YAP/TAZ promote SC proliferation. Additional efforts to confirm this notion and to understand the presumably axon-dependent mechanisms that tightly regulate nuclear levels, thus transcriptional activity, of YAP/TAZ in SCs may generate new strategies for peripheral nerve repair.

Materials and methods

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional information
Strain, strain background (Mus musculus)C57Bl/6Jackson LaboratoryStock #: 000664; RRID:IMSR JAX:000664
Genetic reagent (M. musculus)Plp1-Cre-ERT2MGI:2663093(Leone et al., 2003)
Antibodyanti-Yap/Taz (rabbit monoclonal)Cell Signaling TechnologyD24E4, #8418
RRID:AB_10950494
IHC 1:200
Western 1:1000
Antibodyanti-SCG10 (rabbit monoclonal)Novus BiologicalsNBP1-49461
RRID:AB_10011569
IHC 1:5000
Antibodyanti-Yap (rabbit monoclonal)Cell Signaling TechnologyD8H1X, #14074
RRID:AB_2650491
IHC 1:200
Antibodyanti-Sox10 (goat polyclonal)R and D Systems#AF-2864
RRID:AB_442208
IHC 1:100
Antibodyanti-Sox10 (rabbit monoclonal)AbcamEPR4007, #ab155279
RRID:AB_2650603
IHC 1:250
Antibodyanti-Egr2 (rabbit polyclonal)Professor Dies Meijer, University of EdinburghIHC 1:4000
Antibodyanti-Oct6 (rabbit monoclonal)AbcamEP5421, #ab126746
RRID:AB_11130256
WB 1:1000
Antibodyanti-Oct6 (rabbit polyclonal)Abcam#ab31766
RRID:AB_776899
IHC 1:800
Antibodyanti-c-Jun (mouse monoclonal)BD Transduction Laboratories#610326
RRID:AB_397716
IHC 1:500
Antibodyanti-c-Jun (rabbit monoclonal)Cell Signaling Technology60A8, #9165
RRID:AB_2130165
WB 1:1000
Antibodyanti-pS63-c-Jun (rabbit polyclonal)Cell Signaling Technology#9261
RRID:AB_2130162
IHC 1:100
Antibodyanti-Ki67 (rabbit polyclonal)Abcam#ab15580
RRID:AB_443209
IHC 1:200
Antibodyanti-p75NGFR (goat polyclonal)Neuromics#GT15057
RRID:AB_2737189
IHC 1:400
Antibodyanti-Tubulin β3 (rabbit polyclonal)Biolegend#802001
RRID:AB_2564645
IHC 1:1000
AntibodyIRDye-680 (goat anti-mouse)LI-COR#926–32220
RRID:AB_621840
WB 1:15,000
AntibodyHRP-Goat anti-mouse secondary antibodyJackson Immunoresearch#715-035-150
RRID:AB_2340770
WB 1:12,000
AntibodyHRP-Goat anti-rabbit secondary antibodyJackson Immunoresearch#115-055-062
RRID:AB_2338533
WB 1:12,000
Chemical compound, drugAraldite 6005EMS#10920
Chemical compound, drugDDSAEMS#13710
Chemical compound, drugDBPEMS#13101
Chemical compound, drugBDMAEMS#11400–25
OtherCoated grids (100 mesh)EMS#FF100-Cu
Chemical compound, drugOsmium tetroxide (4% solution)EMS#19170
Chemical compound, drugLead nitrateEMS#17900
Chemical compound, drugSodium citrateEMS#21140
Chemical compound, drugUranyl acetateEMS#22400
Chemical compound, drugSodium borateEMS#21130
Chemical compound, drugToluidine blueEMS#22050
Chemical compound, drugParaformaldehydeSigma-Aldrich#158127
Commercial assay or kitClick-It EdU Alexa Fluor 594 kitThermoFisher Scientific#C10339
Chemical compound, drugEdUThermoFisher Scientific#E10187
Chemical compound, drugTamoxifenSigma-Aldrich#T5648
OtherDAPI stainInvitrogen#D1306IHC 1:250
AntibodyAlexa 488, 568 or 647 secondariesJackson ImmunoresearchIHC 1:250 to 1:1000
Software, algorithmImage Studio LiteLI-COR, Inc
Software, algorithmPrismGraphPad Software, Inc
Software, algorithmStataStataCorp LPMann-Whitney test

Animals

All surgical procedures and animal maintenance complied with the National Institute of Health guidelines regarding the care and use of experimental animals and were approved by the Institutional Animal Care and Use Committee of Temple University, Philadelphia, PA, USA (Protocol 4920). Both male and female mice were used in all experiments, and were maintained on the C57BL/6 background. Plp1-creERT2; Yapfl/fl; Tazfl/fl, Plp1-creERT2; Yap+/+; Tazfl/fl, Mpz-cre;Yapfl/fl and Mpz-cre; Tazfl/fl mice used in this study were generated and genotyped as described previously (Grove et al., 2017). C57BL/6 mice were used for immunohistochemical analysis of YAP/TAZ.

Tamoxifen administration

Request a detailed protocol

Tamoxifen was injected into 6–8 week old Yap/Taz iDKO or Taz iKO mice as previously described (Grove and Brophy, 2014). A 10 mg/ml solution of tamoxifen was made in 10:1 sunflower oil: 100% ethanol. This solution was injected intraperoneally at a concentration of 0.2 mg/g body weight. Injection was once daily for 5 days, followed by a 2 day break, then once daily for 5 consecutive days.

Nerve crush or transection

Request a detailed protocol

Sciatic nerves of right hindlimbs were crushed or transected 24 hr after the final tamoxifen injection, using standard protocols (Son and Thompson, 1995). Briefly, a small skin incision was made in the posterior thigh and calf of the animals anesthetized by isoflurane. For crush, the sciatic nerve was crushed with a fine forceps (#5) for 10 s (3X) adjacent to the sciatic notch. The crush site was marked using charcoal-coated forceps, and the wound was closed. For transection, the exposed sciatic nerve was ligated at two directly adjacent sites, then cut with iridectomy scissors between the ligated sites. Ligated proximal and distal nerve endings were then sewn to adjacent muscle to prevent regeneration of axons from the proximal to distal nerve stumps. To identify proliferating Schwann cells, we intraperitoneally injected EdU (80 μg/g) eighty minutes before killing mice, as previously described (Grove et al., 2017).

Western blotting

Request a detailed protocol

Mice were perfused with PBS, sciatic nerves removed, and epineurium and perineurium carefully stripped from the nerves. Western blotting followed the same procedure described previously (Grove et al., 2017), except for IRDye 680RD goat anti-mouse IgG (LiCor #926–68070; 1:5,000). Image Studio Lite (LI-COR Biosciences) was used for quantifying protein expression.

Immunohistochemistry

Request a detailed protocol

Sciatic nerves were removed, and immediately fixed in 4% paraformaldehyde in PBS for 1 hr on ice. Nerves were washed 3 times in PBS, then stored in 15% sucrose in PBS overnight at 4°C for cryoprotection. Nerves were frozen-embedded in cryoprotectant medium (Thermo Fisher Scientific, Waltham, MA) in isomethylbutane at −80°C. 7–10 μm sections from the nerves were cut using a cryostat (Leica Microsystems, Germany) and collected directly onto glass slides. For immunolabeling, nerve sections were rehydrated in PBS, permeabilized in 0.5% Triton/PBS for 20 min, washed with PBS, then blocked in 2% bovine serum albumin (BSA) in PBS for 1 hr. Sections were incubated with primary antibodies in blocking buffer overnight at 4°C in a hydrated chamber, washed with PBS, and incubated with secondary antibodies in blocking buffer for 2 hr at room temperature. Sections were washed with PBS, stained with DAPI for 10 min, and mounted with Vectashield mounting medium (Vector Labs, Burlingame, CA). Nerve sections were incubated with antibodies previously described (Grove et al., 2017), except for the following: rabbit anti-Krox20 (kind gift from Professor Dies Meijer, Edinburgh, UK; 1:4000), rabbit anti-Yap (Cell Signaling #14074; 1:200), rabbit anti-SCG10 (Novus Biologicals #49461; 1:5000), goat anti-Sox10 (Santa Cruz #sc-17342; 1:200), goat anti-Sox10 (R and D Systems #AF-2864; 1:100), goat anti-p75 (Neuromics #GT15057; 1:400), rabbit anti-Ki67 (Abcam #ab15580; 1:1000), mouse anti-Tubulin β3 (clone Tuj1, Covance #MMS-435P; 1:1000), mouse anti-cJun (BD Biosciences #610326; 1:500), rabbit anti-cJun (CST #9165; 1:500), rabbit anti-phospho-cJun (CST #9261; 1:100).

Electron microscopy, histology and morphometry

Request a detailed protocol

Sciatic nerves were removed and immediately fixed in EM buffer, as previously described (Grove et al., 2017). After nerve crush or transection, a 5 mm piece of the nerve was taken immediately distal to the injury site. The proximal end of the section was nicked with a razor blade for orientation during embedding. Fixation was for 2 hr at room temperature, followed by overnight at 4°C, with rotation. Post-fixation processing, embedding, cutting, staining and image capture were as previously described. For crushed or transected nerves, 500 nm semi-thin and 70 nm ultra-thin transverse sections were cut from the segment 5 mm distal to the crush/transection site.

For analysis of axon regeneration and remyelination, 7500x TEM sections were examined. This magnification allowed unambiguous identification of basal lamina tubes through which axons regenerate. Multiple non-overlapping images were taken for each section, such that all regions of each section were sampled. Image J was used for image analysis. After counting the total number of basal lamina (BL) tubes per image, we next counted the number of BL tubes in the following categories: contains no axon(s); contains axon(s); contains at least one axon >1 μm in diameter; contains a single axon >1 μm in diameter; contains a myelinated axon. This procedure enabled us to calculate the percentage of BL tubes in each category. Using an ImageJ G-ratio calculator plug-in, G ratios for each genotype were calculated in two different ways: (1) All single large axons were counted, whether or not they were myelinated; (2) Only myelinated axons were counted.

Data analysis

Request a detailed protocol

In each experiment, data collection and analysis were performed identically, regardless of mouse genotype. Data are presented as mean + / - SD. Statistical analysis was done using the two-sample Mann-Whitney test for two-group comparisons and analysis of variance (ANOVA) with Tukey’s test for multiple comparisons, according to the number of samples and the analysis of mice at multiple ages. Sample sizes were similar to those employed in the field and are indicated in the main text, methods or figure legends. A p-value of 0.05 or less was considered statistically significant.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
    How schwann cells sort axons: new concepts
    1. ML Feltri
    2. Y Poitelon
    3. SC Previtali
    (2016)
    The Neuroscientist : A Review Journal Bringing Neurobiology, Neurology and Psychiatry 22:252–265.
    https://doi.org/10.1177/1073858415572361
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50

Decision letter

  1. Anna Akhmanova
    Senior and Reviewing Editor; Utrecht University, Netherlands
  2. Michael Wehr
    Reviewer; Universität München, Germany

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

This follow-up study on the authors' previous work examines the effects of transcriptional activity of YAP/TAZ in Schwann cell response to peripheral nerve injury. Using in vivo analyses of mutant mice, the authors show that YAP/TAZ function is not essential for Schwann cell proliferation and transdifferentiation into repair cells but is required for remyelination. This comprehensive work advances our understanding of regeneration of peripheral nervous system.

Decision letter after peer review:

Thank you for submitting your article "Axon-dependent expression of YAP/TAZ mediates Schwann cell remyelination but not proliferation after nerve injury" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Anna Akhmanova as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Michael Wehr (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.

The reviewers liked your work, in which you sought to determine the role of YAP/TAZ after nerve injury. This work is a follow-up of an earlier paper in which the team had assessed the role of these transcription factors in developing Schwann cells and myelination.

While YAP/TAZ deletion in Schwann cells does not affect Schwann cells trans-differentiation into repair Schwann cells (nor axonal regeneration) following PNS injury, the authors show here that YAP and TAZ control the extent of remyelination after injury. YAP and TAZ disappear in denervated Schwann cells, but re-appear in these cells when axons regenerate. In addition, the authors report an effect on Krox 20 expression in mutant mice but not of several other factors, such as c-jun, p75NTR, Oct6. Mutant nerves fail to form basal lamina bundles during regeneration. The findings are interesting and potentially significant to the field of PNS injury

The work has been thoroughly performed and the data presented support the overall assessment. Nevertheless, there are a number of issues that need to be addressed.

Major points:

1) Figure 1

Figure 1B: It is not clear whether denervation results in loss of YAP/TAZ expression or prevents the nuclear localization (thus, the transcriptionally activity) in Schwann cells. If it is the latter, YAP/TAZ is expected to translocate to the cytoplasm following injury, which does not seem to be the case. Accordingly, the authors state that both nuclear and cytoplasmic YAP/TAZ are down regulated in Schwann cells. However, the sentence "SCs are critically dependent on axons for the transcriptional activity of YAP/TAZ" could be misleading and should be rephrased.

Figure 1D: the authors mention that YAP is weakly expressed 24dpi, nevertheless the extent of expression is comparable to that observed at 6dpi. Please rephrase as this assay is not quantitative.

Figure 1 E and F show that disappearance and re-appearance of YAP/TAZ correlates with the timing of axonal degeneration and regeneration, respectively. However, the authors do not show that YAP/TAZ expression depends on the axon. To make the claim, one should look for direct evidence and causality.

2) The conclusion that YAP expression critically depends on axons (subsection “YAP/TAZ expression in Schwann cells is axon-dependent”) appears too strong as at 6dpi axons are starting to regenerate. This concept is also extensively presented in the Discussion where it is claimed that YAP nuclear expression "is highly dependent on axons" (Discussion paragraph three). While this is likely true 24 dpi, at 6 dpi the axons are still (actively) degenerating.

3) One of the main observations in iDKO is the defect in basal lamina (BL) tube formation. Can one exclude that this defect is due to a different deposition of BL components between WT and iDKO mice? Can the authors perform a western blotting analyses to determine its composition?.

4) The reduction in Krox 20 expression could alternatively be a consequence rather than the cause of impaired remyelination. Please discuss and soften conclusions.

5) In their discussion the authors posit that the disappearance of YAP/TAZ from denervated nerves provides support for a role in myelin maintenance, unfortunately without data corroborating this. It is also possible that YAP/TAZ dowregulation is merely due to the ongoing de-differentiation occurring in injured Schwann cells, which is an essential pre-requisite to acquire a repair phenotype (Discussion paragraph one).

6) Based on the lack of expression of YAP/TAZ at 3dpi, one cannot conclude SC-axon interactions are essential to maintain YAP/TAZ expression in adult SCs. Denervated Schwann cells have a completely different phenotype from those observed in adult nerves (Discussion paragraph three).

7) In Discussion paragraph four, the authors propose that "YAP/TAZ powerfully stimulate developing and adult SCs to proliferate at all stages". This conclusion is not directly supported by data. The authors point out that in a recent study (Mindos et al., 2017) YAP and TAZ are critically regulated by Merlin in Schwann cells (SCs), and inactivation of Merlin leads to higher levels of YAP and TAZ preventing axon regeneration and remyelination. However, it remains unclear how YAP (in Taz iKO nerves) can myelinate regenerated axons. Would removal of Merlin in Taz iKO nerves also prevent myelination? An experiment that identifies these mechanisms would help.

8) The authors should check the statistics they applied. A t-test assumes that samples are normally distributed. This is not likely the case for 3 replicates. Use the Mann-Whitney test as alternative. Western blots should also be quantified and for all quantifications (imaging, Western blotting) individual data points should be shown.

9) Figure 4: Does the loss of YAP/TAZ result in a delay of axonal re-generation? Mindos et al., 2017, had shown in mice hat regenerating axons could reach up to 13-14 mm distal to the injury site within 7 days of nerve crush. The present study accessed regeneration 12-14 days after injury but only within the 5-10 mm distal region, a paradigm unlikely to detect a delay in axonal regeneration if present. To address this issue, regeneration should be analyzed earlier and ideally along its time course.

10) Similarly Figures 6, 7: Nerve regeneration/remyelination was accessed 12-13 dpi, a couple of days before death of the mice is expected. As stated in subsection “SCs lacking YAP/TAZ convert to repair SCs and support axon regeneration”, iDKO mice die about 14 days after tamoxifen injection (also stated in Grove et al., 2017). At the time point analyzed, even uninjured iDKO mice exhibit clear PNS nerve defects (shivering, decrease in nerve conduction velocity, demyelination etc) as well as ataxia and impaired breathing (Grove et al., 2017). This complicates interpretation of the data on remyelination as the failure to re-build myelin could be due to a secondary effect stemming from the poor physiological condition of the mice. To prove that the remyelination failure is indeed due to the loss of YAP/TAZ, one would like to see a rescue experiment that restores the YAP/TAZ function in the iDKO Schwann cells. Unless such experiments are provided, the conclusion and discussion must be carefully rephrased.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Axon-dependent expression of YAP/TAZ mediates Schwann cell remyelination but not proliferation after nerve injury" for further consideration by eLife. Your revised article has been evaluated by Anna Akhmanova (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below. The reviewers agreed that their remaining concerns can be addressed without performing additional experiments, by textual changes and clarifications and by providing source data.

Reviewer #1:

The revised version of the manuscript has significantly improved as compared to the original study. Nevertheless, the authors have not performed the requested additional experiments, which are instead argued in the rebuttal letter.

While it is clear from the presented data that p-YAP relocalizes in cytoplasm (Figure 2D), in Figure 1B the authors show and state that " the dramatic down- and up- regulation of YAP/TAZ concurrent with axon degeneration and regeneration, respectively, suggest that SCs are dependent on axons from YAP/TAZ nuclear localization". Given the experimental data it would be more correct talking about expression rather than localization. The latter indeed assumes a shuttle between in an out of the nucleus, which is valid for p-YAP and also in that case only partially.

Reviewer #2:

Grove et al. provide a detailed point-to-point response that satisfactorily discusses all issues raised.

However, this reviewer still has the following concerns related to the quantification of WB data (see below). Once clarified, the manuscript can be granted for publication.

Figure 1C, Figure 1—figure supplement 2:

For the quantification, 4 data points are shown per condition. Supplemental data show 2 uncropped blots that are 2 different exposures from the very same gel as presented in 1C (as the authors also specify). However, 3 independently performed blots that should show independently processed samples and were used for the quantification are not available. The authors need to provide the correct source data. Also note that Figure 1—figure supplement 2 contains a staining of nerve fibers, but no WB data.

Figure 4—figure supplement 1 and 2:

The quantification implicates two blots that were quantified. However, supplement 2 contains the same blot as presented in supplement 1. The authors need to provide the correct source data.

Figure 8A, Figure 8—figure supplement 2:

The quantification implicates 3 blots that were quantified. However, supplement 2 contains the same blot as presented in Figure 8A. Again, the authors need to provide the correct source data for all blots quantified.

Reviewer #3:

This is a revised manuscript which investigates the role of YAP/TAZ in mediating Schwann cell response to injury and PNS regeneration. The authors have addressed reviewers concerns sufficiently, by either conducting additional experiments or in writing. Overall, data from the study are solid and the findings are potentially significant to the field of PNS injury and regeneration. I only have one concern, which can be addressed in writing by the authors.

Figure 1 and Figure 1—figure supplement 1 (paragraph four subsection “YAP/TAZ expression in Schwann cells is axon-dependent”). I am still confused about the discrepancy between the WB and IHC results on YAP expression in denervated nerves. When detected WB, YAP expression continues to decrease after nerve crush and does not recover, whereas when assessed by IHC, YAP expression returns at 6 and 12D. The authors claim that the WB result reflects YAP expression in other cell types (I have to say, I do not agree with the explanation). However, they also make a point in the Discussion that WB was done using lysates prepared from nerves that had been de-sheathed (epi- and perineurium removed) to ensure enrichment of SC-derived proteins.

Since most of the YAP is nuclear at 6 and 12D, shown by the IHC, is it possible that the lysis buffer used for WB does not recover nuclear YAP in the Schwann cells, thus the WB only reflects the cytoplasmic YAP?

https://doi.org/10.7554/eLife.50138.sa1

Author response

Major points:

1) Figure 1

Figure 1B: It is not clear whether denervation results in loss of YAP/TAZ expression or prevents the nuclear localization (thus, the transcriptionally activity) in Schwann cells. If it is the latter, YAP/TAZ is expected to translocate to the cytoplasm following injury, which does not seem to be the case. Accordingly, the authors state that both nuclear and cytoplasmic YAP/TAZ are down regulated in Schwann cells. However, the sentence "SCs are critically dependent on axons for the transcriptional activity of YAP/TAZ" could be misleading and should be rephrased.

Figure 1D: the authors mention that YAP is weakly expressed 24dpi, nevertheless the extent of expression is comparable to that observed at 6dpi. Please rephrase as this assay is not quantitative.

Figure 1 E and F show that disappearance and re-appearance of YAP/TAZ correlates with the timing of axonal degeneration and regeneration, respectively. However, the authors do not show that YAP/TAZ expression depends on the axon. To make the claim, one should look for direct evidence and causality.

Figure 1B (now Figure 2B): We have added Figure 2D, which shows that cytoplasmic expression of YAP (i.e., pYAP) returns in Schwann cells (SCs) of the crushed nerve, but not in those of the transected nerve in which axon regeneration into the distal stump is prevented (Figure 2C). This data are additional evidence that axons are necessary for the recovery of cytoplasmic expression of YAP/TAZ in SCs. We have also amended our statements on axon-dependence of YAP/TAZ expression.

Figure 1—figure supplement 1D (now Figure 1—figure supplement 1B): We have removed the confusing statement and just mention here that “YAP is upregulated in many SC nuclei at or after 6 dpi…”.

Figure 1E and F (now Figure 1B and C): Please note that we do NOT claim that SCs depend on “axon contact” for YAP/TAZ expression. We believe that this is likely to be the case (see below #2 and #6 responses) but we agree with the reviewer that only correlative data support it. We claim that they depend on “axons” (i.e., presence of axons implicating neural regulation) mainly based on the nerve transection experiment (previous Figure 1B and C, now Figure 2). Whereas crushing nerve permits regeneration of axons into the distal nerve stump, completely cutting and tying nerve (as we did in the nerve transection experiment) prevents axons from regenerating into distal nerve stump. As the distal stump of transected/tied nerves will not have axons, comparing protein expression in crushed versus transected/tied nerve has been used to examine axonal dependence or neural regulation in vivo. Our finding that YAP/TAZ disappear in SCs upon axon degeneration in both crushed and transected nerves, that they reappear in crushed nerve concomitant with regenerating axons, but not in transected nerve lacking axons therefore provides strong evidence for the dependence of YAP/TAZ expression in SCs on axons. Whether this neural regulation is via direct axon-SC contact or releasable factors or is mediated by another cell type awaits additional study.

2) The conclusion that YAP expression critically depends on axons (subsection “YAP/TAZ expression in Schwann cells is axon-dependent”) appears too strong as at 6dpi axons are starting to regenerate. This concept is also extensively presented in the Discussion where it is claimed that YAP nuclear expression "is highly dependent on axons" (Discussion paragraph three). While this is likely true 24 dpi, at 6 dpi the axons are still (actively) degenerating.

It is true that axon degeneration can be active at 6 dpi or even later. It is also important to note that, in the nerve crush injury model, new axons form and penetrate the crush site within 1-2 days after injury and extend further along the distal nerve stump. As shown in the newly provided Figure 6 and in many other studies, regenerating axons can reach up to ~4 mm distal to the crush site at 3 dpi. These axons then keep regenerating within the basal lamina (BL) tubes at the speed of 1-4 mm/day, although the debris of degenerating axons and myelin are not yet completely removed. We observed regenerating axons at 3 dpi (e.g., Figure 1B; zoomed area of 3D-dstl) but many more (and thicker, large-diameter) axons at 6 dpi which made contacts with SCs (Figure 1B; zoomed area of 6D-dstl). These data therefore support our conclusion and further suggest that regenerating axons upregulate YAP/TAZ in SCs. However, we agree with the reviewer that our statements were too strong, and we have modulated the sentences.

3) One of the main observations in iDKO is the defect in basal lamina (BL) tube formation. Can one exclude that this defect is due to a different deposition of BL components between WT and iDKO mice? Can the authors perform a western blotting analyses to determine its composition?.

We assume that the reviewer meant failed remyelination by “…defect in BL formation…”, and that s/he wondered if failed remyelination in iDKO could be due to defective signaling from BL evoked by the loss of YAP/TAZ in SCs. First, it is our understanding that BL is important but axons and SCs also provide essential components for optimal remyelination. The complete, rather than partial or delayed, loss of remyelination that we observed in iDKO is therefore likely to be due to inability of SCs in iDKO to coordinate signals from axons and SCs, as well as those from BL. Second, we consider it unlikely that substantial changes can be made in BL tubes in adult iDKO mice, as they are already established and completely normal before tamoxifen injection. Considering the marked structural and molecular stability of BL tubes in mature nerves, we anticipate modest if any changes in BL composition during the relatively short time between tamoxifen injection and terminal end points. In support of this notion, axons regenerate normally within BL tubes in iDKO (Figure 5). Axon regeneration should have been attenuated if there were substantial changes in BL associated growth-promoting molecules such as laminin. Laminin is a main BL component also known to promote myelination. Thus, we believe it reasonable to predict that SCs fail to remyelinate although a substantial amount of laminin is present in BL of iDKO.

4) The reduction in Krox 20 expression could alternatively be a consequence rather than the cause of impaired remyelination. Please discuss and soften conclusions.

We appreciate the comment and have amended our statements and discussed as suggested. Krox 20 is widely accepted as the key transcription factor (TF) promoting peripheral myelination. This is largely believed to be the case after injury, although its role in remyelination has never been explicitly demonstrated. Other pathways can also promote upregulation of certain myelin proteins and lipids independently of Krox 20, and numerous other factors, both positive and negative regulators, mediate peripheral myelination. Therefore, it is conceivable that the lack of Krox 20 in iDKO is a consequence rather than a cause of impaired myelination. However, several considerations make it highly unlikely. First, we and others have shown that YAP/TAZ-TEAD1 complex directly binds to the cis-acting regulatory sequence of Krox 20 (i.e., MSE) to upregulate Krox 20 during developmental myelination. Second, we showed that YAP/TAZ-TEAD1 also bind to Krox 20 MSE in mature nerve, suggesting direct regulation of Krox 20 by YAP/TAZ in mature SCs. Third, Oct 6, which induces upregulation of Krox 20, together with other TFs, is upregulated in iDKO after nerve injury, as in WT (Figure 4H, 9G). This finding suggests that remyelination in iDKO is blocked at the step of Krox 20 upregulation. Indeed, we and others have shown that iDKO SCs are arrested at the promyelination stage during development, and similarly after nerve injury, we found that iDKO SCs are capable of wrapping around individual axons but fail to upregulate Krox 20 and initiate myelin formation (Figure 5E, 7E).

5) In their discussion the authors posit that the disappearance of YAP/TAZ from denervated nerves provides support for a role in myelin maintenance, unfortunately without data corroborating this. It is also possible that YAP/TAZ dowregulation is merely due to the ongoing de-differentiation occurring in injured Schwann cells, which is an essential pre-requisite to acquire a repair phenotype (Discussion paragraph one).

We reduced and clarified our discussion of the “implication” of the present remyelination study for the role of YAP/TAZ in myelin maintenance. Due to conflicting reports among different groups, it remains controversial whether YAP/TAZ are crucial also for maintaining myelin in mature nerve. This is a very important issue in the field, and intimately associated with demyelination and remyelination of mature nerve after injury. Therefore, we believe that our discussion of the “implication” of the present study is sufficiently justified and does not require direct supporting data. We also discuss prior studies that support the notion that YAP/TAZ downregulation may be actively regulated, rather than passive due to ongoing de-differentiation. These findings include that Krox 20 mRNA is not downregulated until 3 days after injury (Arthur-Farraj et al., 2017), when we observed YAP/TAZ protein downregulation. This timing suggests that YAP/TAZ protein downregulation leads to Krox 20 mRNA downregulation, suppressing expression of myelin proteins in de-differentiating SCs. Complete loss of both nuclear and cytoplasmic YAP/TAZ could therefore implicate possible presence of active regulatory mechanisms that completely inactivate YAP/TAZ after injury. In support of this notion, in mutant mice lacking Merlin, YAP is abnormally upregulated in SCs after nerve injury, impairing SC de-differentiation (Mindos et al., 2017). This YAP upregulation suggests that one role of Merlin is to downregulate YAP in SCs after injury and that YAP/TAZ expression in SCs is likely under active regulation in both intact and injured mature nerves.

6) Based on the lack of expression of YAP/TAZ at 3dpi, one cannot conclude SC-axon interactions are essential to maintain YAP/TAZ expression in adult SCs. Denervated Schwann cells have a completely different phenotype from those observed in adult nerves (Discussion paragraph three).

Here again we meant to discuss the implications of the remyelination data for myelin maintenance, not as a conclusion. We have now softened and clarified the statement so that readers do not misinterpret us. As mentioned in our response to #2 comment, we claim that SCs depend on axons for YAP/TAZ expression after injury. We now provide images of an enlarged area of crushed and transected/tied nerves (Figure 1B, Figure 2B, 2C, 2D) to better demonstrate that loss and recovery of YAP/TAP expression in SCs are spatiotemporally correlated with loss and recovery of SC-axon contact. It is particularly interesting that a few thin axons make contacts with SCs as early as 3 dpi but that YAP/TAZ are not upregulated, whereas there is obvious upregulation at 6 dpi when many of these axons become thicker (i.e., >1um or large enough to be myelinated). These observations are consistent with our earlier finding that YAP/TAZ are expressed only in myelinating, but not in non-myelinating, SCs (Grove et al., 2017), and further support the notion that the absence and presence of axons determine YAP/TAZ down- and upregulation in SCs after injury, respectively. We believe that this additional data further justify our suggestion in the discussion that the same mechanism may maintain YAP/TAZ expression in adult SCs in intact nerve.

It is also noteworthy that, despite drastic phenotype differences, both innervated and denervated SCs and their processes reside within basal lamina (BL) tubes. It is also well established that axons degenerate and regenerate within BL tubes, losing or regaining contacts with SCs and SC processes that tightly fill the BL tubes and comprise the bands of Büngner. Therefore, although the confocal microscopy that we used in the present study does not provide sufficient resolution to demonstrate direct cell-cell contact between SCs and the fluorescently labeled axons, the presence of axons closely adjacent to the SC nuclei of interest strongly suggests that these SCs have made direct contact with axons.

7) In Discussion paragraph four, the authors propose that "YAP/TAZ powerfully stimulate developing and adult SCs to proliferate at all stages". This conclusion is not directly supported by data. The authors point out that in a recent study (Mindos et al., 2017) YAP and TAZ are critically regulated by Merlin in Schwann cells (SCs), and inactivation of Merlin leads to higher levels of YAP and TAZ preventing axon regeneration and remyelination. However, it remains unclear how YAP (in Taz iKO nerves) can myelinate regenerated axons. Would removal of Merlin in Taz iKO nerves also prevent myelination? An experiment that identifies these mechanisms would help.

We found that SCs do not require YAP/TAZ to proliferate after nerve injury. This finding was quite unexpected because YAP/TAZ are crucial for the proliferation of developing SCs (Grove et al., 2017). In our discussion of this interesting finding, we suggested that the finding was unlikely to indicate that YAP/TAZ are unable to stimulate proliferation of “mature” SCs, which was what we meant by our statement that "YAP/TAZ…stimulate…adult SCs to proliferate…". We referred to two recent publications from other groups which support the statement, and which reported increased proliferation of mature SCs in response to abnormally high activation of YAP in injured nerve in the absence of Merlin (Mindos et al., 2017) or of YAP/TAZ in intact nerve in the absence of LATS1/2 (Wu et al., 2018). We think that these reports provide sufficient rationale for our statement and discussion.

It is worth noting that both reports found increased YAP-, or YAP/TAZ-dependent SC proliferation when the upstream regulators of YAP/TAZ (Merlin, LATS1/2) were experimentally dysregulated, whereas we observed YAP/TAZ-independent proliferation of SCs when these signals are intact and properly regulating spatiotemporal expression of YAP/TAZ after injury. As we mentioned in #5 response, we postulate that active regulatory mechanism(s) remove YAT/TAZ from SCs to help dedifferentiation upon axon degeneration after injury, and then upregulate them upon axon regeneration to induce remyelination. We also postulate that the regulatory mechanism(s) prevent redifferentiating SCs that contact large regenerating axons from excessively upregulating YAP/TAZ, because excessive activation of YAP/TAZ may stimulate SC proliferation rather than differentiation, as in Merlin- or LATS1/2-null mutants.

Consistent with this notion, we also found that YAP expression remains at a normal level in Taz iKO and that YAP alone promote myelination in the absence of TAZ, although YAP and TAZ together induce more efficient remyelination. We are quite confident to conclude that YAP and TAZ are functionally redundant in remyelination. We do not know why the reviewer questions our conclusion, but speculate that it is related to the report of Mindos et al., 2017. These investigators examined nerve injury in the absence of Merlin and raised the possibility that YAP and TAZ may have different functions, which contradicts our conclusion. This report also included some other discrepancies with our observations, such as TAZ expression in denervated SCs. We are very interested to know if Merlin indeed regulates YAP and TAZ expression differently, the mechanisms by which Merlin regulates YAP/TAZ, and how its loss impacts spatiotemporal expression and thus the activity of YAP/TAZ. This information will be of great interest, but, in our opinion, generation and in-depth analysis of Merlin and Taz double mutants are beyond the scope of the present work.

8) The authors should check the statistics they applied. A t-test assumes that samples are normally distributed. This is not likely the case for 3 replicates. Use the Mann-Whitney test as alternative. Western blots should also be quantified and for all quantifications (imaging, Western blotting) individual data points should be shown.

As suggested, we used the Mann-Whitney U test, quantified additional Western blots (Figure 4—figure supplement 1, Figure 8A) and indicate individual data points on all graphs.

9) Figure 4: Does the loss of YAP/TAZ result in a delay of axonal re-generation? Mindos et al., 2017, had shown in mice hat regenerating axons could reach up to 13-14 mm distal to the injury site within 7 days of nerve crush. The present study accessed regeneration 12-14 days after injury but only within the 5-10 mm distal region, a paradigm unlikely to detect a delay in axonal regeneration if present. To address this issue, regeneration should be analyzed earlier and ideally along its time course.

As requested, we analyzed an earlier phase of regeneration by selectively labeling regenerating axons at 3 days after injury, and the result is presented in the new Figure 6. We observed similar numbers of axons that penetrated the injured site and reached ~4mm distal to the crushed site in YAP/TAZ iDKO, as in WT mice. Thus, it is highly unlikely that regeneration is delayed in the absence of YAP/TAZ. We chose 3 days after injury to assess the onset and early phase of regeneration when axotomized distal axons are actively degenerating. In rodents axons regenerate along the distal stump at a speed of up to 4mm/day, and even the most distal muscles are reinnervated in less than 12 days in mice. We believe, therefore, that if we were to find a difference of 1-2 millimeters in how far the longest axons regenerate at times later than 3 dpi in YAP/TAZ iDKO and WT mice, that this difference would not reliably indicate poor regeneration.

10) Similarly Figures 6, 7: Nerve regeneration/remyelination was accessed 12-13 dpi, a couple of days before death of the mice is expected. As stated in subsection “SCs lacking YAP/TAZ convert to repair SCs and support axon regeneration”, iDKO mice die about 14 days after tamoxifen injection (also stated in Grove et al., 2017). At the time point analyzed, even uninjured iDKO mice exhibit clear PNS nerve defects (shivering, decrease in nerve conduction velocity, demyelination etc) as well as ataxia and impaired breathing (Grove et al., 2017). This complicates interpretation of the data on remyelination as the failure to re-build myelin could be due to a secondary effect stemming from the poor physiological condition of the mice. To prove that the remyelination failure is indeed due to the loss of YAP/TAZ, one would like to see a rescue experiment that restores the YAP/TAZ function in the iDKO Schwann cells. Unless such experiments are provided, the conclusion and discussion must be carefully rephrased.

We now discuss why we consider it highly unlikely that the remyelination failure is due to poor physiological condition of the iDKO mice. First, axons regenerate normally in iDKO, which is unlikely if SCs are selectively vulnerable to poor physiological condition. Furthermore, iDKO SCs proliferate and trans-differentiate to repair-SCs normally. iDKO SCs also downregulate c-Jun to prepare for remyelination, whereas they maintain a higher level of Oct 6 than WT SCs, consistent with the failure of iDKO SCs to upregulate Krox 20 and MBP. Second, iDKO SCs wrap around individual axons, but fail to myelinate them, indicating that they proceed to the promyelination stage but no further. Therefore, one would have to postulate that the poor physiological condition of iDKO mice has a very specific effect on a particular remyelination stage, which we find unlikely.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Reviewer #1:

The revised version of the manuscript has significantly improved as compared to the original study. Nevertheless, the authors have not performed the requested additional experiments, which are instead argued in the rebuttal letter.

While it is clear from the presented data that p-YAP relocalizes in cytoplasm (Figure 2D), in Figure 1B the authors show and state that " the dramatic down- and up- regulation of YAP/TAZ concurrent with axon degeneration and regeneration, respectively, suggest that SCs are dependent on axons from YAP/TAZ nuclear localization". Given the experimental data it would be more correct talking about expression rather than localization. The latter indeed assumes a shuttle between in an out of the nucleus, which is valid for p-YAP and also in that case only partially.

We agree with the reviewer and modified the statement by changing “localization” to “expression”.

Reviewer #2:

Grove et al. provide a detailed point-to-point response that satisfactorily discusses all issues raised.

However, this reviewer still has the following concerns related to the quantification of WB data (see below). Once clarified, the manuscript can be granted for publication.

Figure 1C, Figure 1—figure supplement 2:

For the quantification, 4 data points are shown per condition. supplemental data show 2 uncropped blots that are 2 different exposures from the very same gel as presented in 1C (as the authors also specify). However, 3 independently performed blots that should show independently processed samples and were used for the quantification are not available. The authors need to provide the correct source data. Also note that Figure 1—figure supplement 2 contains a staining of nerve fibers, but no WB data.

There seems to have been confusion in supplement figure numbers. It was Figure 1—figure supplement 1 where we previously presented WB data (i.e., 2 exposures of the same gel in 1C). Figure 1—figure supplement 2 was meant to contain no WB data. The figure that the reviewer requests that we amend is Figure 1—figure supplement 1, not Figure 1—figure supplement 2.

As requested, Figure1—figure supplement 1 now shows additional, uncropped Western Blots from all the mice used for quantification at 4 different time points. They are from independently processed samples from each mouse: 3 mice (1 dpi, 6 dpi), 6 mice (3 dpi) and 4 mice (12 dpi).

Figure 4—figure supplement 1 and 2:

The quantification implicates two blots that were quantified. However, supplement 2 contains the same blot as presented in supplement 1. The authors need to provide the correct source data.

As requested, Figure 4—figure supplement 2 now show uncropped blots from two WT and two iDKO mice, independently processed for quantification.

Figure 8A, Figure 8—figure supplement 2:

The quantification implicates 3 blots that were quantified. However, supplement 2 contains the same blot as presented in Figure 8A. Again, the authors need to provide the correct source data for all blots quantified.

As requested, Figure 8—figure supplement 2 now shows uncropped blots of 3 WT and 3 Taz iKO mice, independently processed for quantification.

Reviewer #3:

This is a revised manuscript which investigates the role of YAP/TAZ in mediating Schwann cell response to injury and PNS regeneration. The authors have addressed reviewers concerns sufficiently, by either conducting additional experiments or in writing. Overall, data from the study is solid and the findings are potentially significant to the field of PNS injury and regeneration. I only have one concern, which can be addressed in writing by the authors.

Figure 1 and Figure 1—figure supplement 1 (paragraph four subsection “YAP/TAZ expression in Schwann cells is axon-dependent”). I am still confused about the discrepancy between the WB and IHC results on YAP expression in denervated nerves. When detected WB, YAP expression continues to decrease after nerve crush and does not recover, whereas when assessed by IHC, YAP expression returns at 6 and 12D. The authors claim that the WB result reflects YAP expression in other cell types (I have to say, I do not agree with the explanation). However, they also make a point in the Discussion that WB was done using lysates prepared from nerves that had been de-sheathed (epi- and perineurium removed) to ensure enrichment of SC-derived proteins.

Since most of the YAP is nuclear at 6 and 12D, shown by the IHC, is it possible that the lysis buffer used for WB does not recover nuclear YAP in the Schwann cells, thus the WB only reflects the cytoplasmic YAP?

For all WBs, sciatic nerves were lysed in RIPA buffer, which is a stringent buffer. Nevertheless, for all antibodies used in WBs, we performed pilot experiments where the sciatic nerve pellet left over after RIPA extraction was dissolved in 3xSDS-PAGE loading buffer (this buffer contains 6% SDS). An equal fraction of the pellet and lysate was run on SDS-PAGE to ensure that all of the protein under study was in the lysate fraction, which was the case with YAP and TAZ. Hence, we are confident that all of the nuclear and cytoplasmic YAP in Schwann cells was in the lysate used for WBs.

IHC revealed numerous SCs exhibiting YAP apparently in their nuclei at 6D or later. However, quantitative analysis was difficult to perform with the sole antibody that specifically recognized YAP (but not TAZ) in IHC. Accordingly, we claim that YAP is upregulated in SCs, but we do not intend to imply that its levels return to normal in SCs. It was indeed our impression that YAP staining intensity may remain weak possibly in quite many SCs even at 24D. We will be interested to explore this possibility, when additional antibodies of high quality are available for YAP or TAZ specific IHC.

https://doi.org/10.7554/eLife.50138.sa2

Article and author information

Author details

  1. Matthew Grove

    1. Shriners Hospitals Pediatric Research Center and Center for Neural Repair and Rehabilitation, Temple University, Philadelphia, United States
    2. Department of Anatomy and Cell Biology, Temple University, Philadelphia, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation
    Competing interests
    No competing interests declared
  2. Hyunkyoung Lee

    1. Shriners Hospitals Pediatric Research Center and Center for Neural Repair and Rehabilitation, Temple University, Philadelphia, United States
    2. Department of Anatomy and Cell Biology, Temple University, Philadelphia, United States
    Contribution
    Data curation, Investigation
    Competing interests
    No competing interests declared
  3. Huaqing Zhao

    Department of Clinical Sciences, Lewis Katz School of Medicine, Temple University, Philadelphia, United States
    Contribution
    Statistical analysis
    Competing interests
    No competing interests declared
  4. Young-Jin Son

    1. Shriners Hospitals Pediatric Research Center and Center for Neural Repair and Rehabilitation, Temple University, Philadelphia, United States
    2. Department of Anatomy and Cell Biology, Temple University, Philadelphia, United States
    Contribution
    Conceptualization, Data curation, Writing - original draft, Writing - review and editing
    For correspondence
    yson@temple.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5725-9775

Funding

National Institute of Neurological Disorders and Stroke (NS105796)

  • Young-Jin Son

Shriners Hospitals for Children (Research award)

  • Young-Jin Son

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Alan Tessler and members of the Son laboratory for critical reading of the manuscript. We thank Dr. Hyukmin Kim for intraperitoneal injection, Dr. Eric Olson for Yap and Taz floxed mice, Drs. Ueli Suter and Kelly Monk for Plp-creERT2 mice. Plp-creERT2 mice were generated by Dr. Ueli Suter using a patented Cre-ERT2 construct developed by Dr. Pierre Chambon at GIE-CERBM. This paper is dedicated to the memory of Dr. Wesley Thompson who revealed the pivotal roles of terminal Schwann cells in forming and restoring nerve-muscle connection.

Ethics

Animal experimentation: All surgical procedures and animal maintenance complied with the National Institute of Health guidelines regarding the care and use of experimental animals and were approved by the Institutional Animal Care and Use Committee of Temple University, Philadelphia, PA, USA. Protocol 4920.

Senior and Reviewing Editor

  1. Anna Akhmanova, Utrecht University, Netherlands

Reviewer

  1. Michael Wehr, Universität München, Germany

Publication history

  1. Received: July 16, 2019
  2. Accepted: May 19, 2020
  3. Accepted Manuscript published: May 21, 2020 (version 1)
  4. Version of Record published: May 29, 2020 (version 2)

Copyright

© 2020, Grove 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.

Metrics

  • 700
    Page views
  • 146
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

  1. Further reading

Further reading

    1. Medicine
    2. Neuroscience
    Mitsuru Shinohara et al.
    Research Article Updated

    Although the ε2 allele of apolipoprotein E (APOE2) benefits longevity, its mechanism is not understood. The protective effects of the APOE2 on Alzheimer’s disease (AD) risk, particularly through their effects on amyloid or tau accumulation, may confound APOE2 effects on longevity. Herein, we showed that the association between APOE2 and longer lifespan persisted irrespective of AD status, including its neuropathology, by analyzing clinical datasets as well as animal models. Notably, APOE2 was associated with preserved activity during aging, which also associated with lifespan. In animal models, distinct apoE isoform levels, where APOE2 has the highest, were correlated with activity levels, while some forms of cholesterol and triglycerides were associated with apoE and activity levels. These results indicate that APOE2 can contribute to longevity independent of AD. Preserved activity would be an early-observable feature of APOE2-mediated longevity, where higher levels of apoE2 and its-associated lipid metabolism might be involved.

    1. Neuroscience
    Maria M Diehl et al.
    Research Article Updated

    The prefrontal cortex (PFC) integrates incoming information to guide our actions. When motivation for food-seeking competes with avoidance of danger, the PFC likely plays a role in selecting the optimal choice. In platform-mediated active avoidance, rats avoid a tone-signaled footshock by stepping onto a nearby platform, delaying access to sucrose pellets. This avoidance requires prelimbic (PL) PFC, basolateral amygdala (BLA), and ventral striatum (VS). We previously showed that inhibitory tone responses of PL neurons correlate with avoidability of shock (Diehl et al., 2018). Here, we optogenetically modulated PL terminals in VS and BLA to identify PL outputs regulating avoidance. Photoactivating PL-VS projections reduced avoidance, whereas photoactivating PL-BLA projections increased avoidance. Moreover, photosilencing PL-BLA or BLA-VS projections reduced avoidance, suggesting that VS receives opposing inputs from PL and BLA. Bidirectional modulation of avoidance by PL projections to VS and BLA enables the animal to make appropriate decisions when faced with competing drives.