1. Cell Biology
  2. Neuroscience
Download icon

Specific labeling of synaptic schwann cells reveals unique cellular and molecular features

  1. Ryan Castro
  2. Thomas Taetzsch
  3. Sydney K Vaughan
  4. Kerilyn Godbe
  5. John Chappell
  6. Robert E Settlage
  7. Gregorio Valdez  Is a corresponding author
  1. Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, United States
  2. Center for Translational Neuroscience, Robert J. and Nancy D. Carney Institute for Brain Science and Brown Institute for Translational Science, Brown University, United States
  3. Neuroscience Graduate Program, Brown University, United States
  4. Fralin Biomedical Research Institute at Virginia Tech Carilion, United States
  5. Department of Advanced Research Computing, Virginia Tech, United States
  6. Department of Neurology, Warren Alpert Medical School of Brown University, United States
Research Article
  • Cited 1
  • Views 1,041
  • Annotations
Cite this article as: eLife 2020;9:e56935 doi: 10.7554/eLife.56935

Abstract

Perisynaptic Schwann cells (PSCs) are specialized, non-myelinating, synaptic glia of the neuromuscular junction (NMJ), that participate in synapse development, function, maintenance, and repair. The study of PSCs has relied on an anatomy-based approach, as the identities of cell-specific PSC molecular markers have remained elusive. This limited approach has precluded our ability to isolate and genetically manipulate PSCs in a cell specific manner. We have identified neuron-glia antigen 2 (NG2) as a unique molecular marker of S100β+ PSCs in skeletal muscle. NG2 is expressed in Schwann cells already associated with the NMJ, indicating that it is a marker of differentiated PSCs. Using a newly generated transgenic mouse in which PSCs are specifically labeled, we show that PSCs have a unique molecular signature that includes genes known to play critical roles in PSCs and synapses. These findings will serve as a springboard for revealing drivers of PSC differentiation and function.

Introduction

The neuromuscular junction (NMJ) is a tripartite synapse comprised of an α-motor neuron (the presynapse), extrafusal muscle fiber (the postsynapse), and specialized synaptic glia called perisynaptic Schwann cells (PSCs) or terminal Schwann cells. Due to its large size and accessibility, extensive research of the NMJ has been essential to the discovery of the fundamental mechanisms that govern synaptic function, including the concepts of neurotransmitter release, quantal transmission, and active zones, among others (Katz and Miledi, 1967; Fatt and Katz, 1952; Sealock et al., 1989; Sobel et al., 1979; Sobel et al., 1977; Sanes and Lichtman, 1999; Darabid et al., 2014; Katz and Miledi, 1966; Robertson, 1956; Changeux et al., 1970; Godfrey et al., 1984; Jennings et al., 1993; Lwebuga-Mukasa et al., 1976; Nitkin et al., 1987; Porter and Froehner, 1983). Likewise, the concept of glia that exist primarily to support synapse function, and thus the realization that synapses are tripartite, has its origins at the NMJ (Robertson, 1956; Couteaux, 1960; Kang et al., 2007; Zuo et al., 2004; Griffin and Thompson, 2008; Boeke, 1949; Heuser and Reese, 1973; Miledi and Slater, 1968; Miledi and Slater, 1970; Peper et al., 1974; Astrow et al., 1994; Astrow et al., 1998; Reynolds and Woolf, 1992; Young et al., 2005). PSCs surround the NMJ where they are closely associated with its pre- and postsynaptic components (Griffin and Thompson, 2008; Ko and Robitaille, 2015; Darabid et al., 2014). In addition to providing trophic support for the NMJ (Griffin and Thompson, 2008; Ko and Robitaille, 2015; Darabid et al., 2014; Reddy et al., 2003), PSCs have been shown to guide motor axon innervation and synaptogenesis (Reddy et al., 2003; Trachtenberg and Thompson, 1997; Koirala et al., 2000; O'Malley et al., 1999; Barik et al., 2016), support compensatory axonal sprouting (Astrow et al., 1994; Reynolds and Woolf, 1992; Son and Thompson, 1995; Love and Thompson, 1998), participate in synaptic pruning (Griffin and Thompson, 2008; Lee et al., 2017; Smith et al., 2013; Darabid et al., 2013), and detect and modulate cholinergic transmission (Ko and Robitaille, 2015; Jahromi et al., 1992; Reist and Smith, 1992; Robitaille, 1995; Robitaille et al., 1997; Rochon et al., 2001).

While great progress has been made in understanding the cellular and physiological characteristics of PSCs, very little is known about the molecular composition of these cells (Ko and Robitaille, 2015). This has been due to the absence of a cell-specific molecular marker with which PSCs can be identified, isolated, and genetically manipulated. This has hindered examinations of the processes of PSC development, differentiation and turnover. Additionally, isolation and targeting of PSCs for interrogation of molecular function in vivo and in vitro has not been possible. Therefore, the discovery of markers specific to PSCs is necessary to advance our understanding of PSCs, and synaptic glia in general, on multiple fronts.

A growing number of molecular markers that recognize subsets of glial cells throughout the nervous system have been identified (Jäkel and Dimou, 2017). Therefore, we explored the possibility that a unique combination of glial cell markers could be used to distinguish PSCs. We have found that PSCs can be identified by the combined expression of the calcium-binding protein B (S100β) (Brockes et al., 1979; Perez and Moore, 1968) and neuron-glia antigen-2 (NG2) (Stallcup, 1981; Bergles et al., 2010) genes. We utilized this unique molecular fingerprint to create a transgenic mouse that enables visualization and isolation of PSCs in a cell specific manner. This genetic model will help overcome obstacles to understanding the cellular and molecular rules that govern PSC function at NMJs during development, following injury, in old age, and in diseases, such as Amyotrophic Lateral Sclerosis (ALS).

Results

To identify unique markers for PSCs, we examined the expression of genes shown to be co-expressed with well-established markers of Schwann cells in a subset of glial cells in the central nervous system in PSCs. We focused on NG2 for the following reasons: 1) it has been found to be co-expressed with S100β, a classical marker of all Schwann cells, in a subset of glial cells in the developing brain (Matthias et al., 2003; Hachem et al., 2005; Vives et al., 2003; Moshrefi-Ravasdjani et al., 2017; Platel et al., 2009); 2) published data show that it is expressed in skeletal muscles and labels pericytes and neural progenitor cells (Birbrair et al., 2013a; Birbrair et al., 2013b). To determine if NG2 is expressed by PSCs, we examined whole-mounted extensor digitorum longus (EDL) muscles from NG2-dsRed mice (Zhu et al., 2008). We observed widespread distribution of NG2-dsRed postive cells in the EDL muscle, including a distinct subset of cells located specifically at the NMJ and with a similar morphology as PSCs (Figure 1C). To determine if NG2 is expressed in PSCs we generated a transgenic mouse line (referred herein as S100β-GFP;NG2-dsRed; Figure 1A) by crossing the NG2-dsRed line with the S100β-GFP mouse line in which the S100β promoter drives expression of GFP in all Schwann cells (Zuo et al., 2004). As expected, in the resulting S100β-GFP;NG2-dsRed double transgenic mouse line, dsRed labeled all NG2 positive cells (referred herein as NG2-dsRed+) and GFP labeled all Schwann cells (referred herein as S100β-GFP+) (Figure 1B–C) in skeletal muscles. However, we found a select subset of glia positive for both S100β-GFP and NG2-dsRed specifically located at the NMJ (yellow cells in Figure 1D). Based on the location and morphology of the cell body and its elaborations, we concluded that PSCs are the only cells expressing both S100β-GFP and NG2-dsRed in skeletal muscles.

Co-expression of S100β and NG2 is unique to PSCs in the EDL muscle.

(A) In order to selectively label PSCs, S100β-GFP and NG2-dsRed transgenic mice were crossed to create S100β-GFP;NG2-dsRed mice. (B–D) Representative images of GFP (B) and dsRed (C) fluorescence in the EDL of S100B-GFP;NG2-dsRed mice. S100β-GFP+ Schwann cells are visible along the motor axon while S100β-GFP+ PSCs are identified by their unique morphology and clustering pattern near the NMJ, visualized here using a fluourescent α-bungarotoxin conjugate (fBTX) to detect nAChRs (blue). Note that PSCs are the only cells expressing both GFP and dsRed (D). At non-synaptic sites, GFP-positive cells do not express dsRed (hollow arrow; B’, C’, D’) and dsRed-positive cells do not express GFP (filled arrow; B’, C’, D’). Scale bar = 50 μm (D), 25 μm (D’), and 10 μm (D’’).

We next evaluated whether the S100β-GFP;NG2-dsRed mouse line serves as a reliable model to study PSCs and their roles at NMJs. In healthy young adult muscle, we observed the same number of PSCs at NMJs in the EDL muscle of S100β-GFP and S100β-GFP;NG2-dsRed mice (Figure 2A–C). The morphology of PSCs also appeared to be indistinguishable between the two transgenic lines. In addition, the morphology of NMJs, as assessed by fragmentation of nicotinic acetylcholine receptor (nAChR) clusters, is unchanged in S100β-GFP;NG2-dsRed mice compared to S100β-GFP and wild type mice (Figure 2A,B,D). Thus, the co-expression of S100β-GFP and NG2-dsRed does not appear to cause apparent deleterious changes on either PSCs or the postsynaptic region revealed by nAChRs. However, it remains possible that co-expression of these markers in PSCs may disrupt the presynapse and biophysical properties of the NMJ. If so, we hypothesize that such changes would be minor given that S100β-GFP;NG2-dsRed mice are outwardly indistinguishable when compared to S100β-GFP and wild type mice.

The NG2-DsRed/S100β-GFP mouse line can be used to reliably identify PSCs in healthy and stressed NMJs.

(A–B) Representative images of NMJs identified by fBTX labeled nAChRs from S100β-GFP (A) and S100β-GFP;NG2-dsRed (B) EDL. (C–D) The co-expression of GFP and dsRed has no discernible negative effects on NMJ fragmentation or PSC number in the EDL muscle of young adult mice. (C) The average number of PSCs per NMJ is unchanged between S100B-GFP mice and S100β-GFP;NG2-dsRed mice. (D) The average number of nAChR fragments per NMJ, as determined by analysis of continuity of fBTX labeled nAChRs, is unchanged between wild-type, S100B-GFP, and S100β-GFP;NG2-dsRed animals. (E–H) PSCs in stressed muscle co-express S100β-GFP and NG2-dsRed. Representative images of NMJs identified by fBTX labeled nAChRs in S100β-GFP;NG2-dsRed mice shows co-expression of S100β-GFP and NG2-dsRed by PSCs in healthy uninjured (E), at 4d (F), and 7d (G) post-fibular nerve crush, and in P120 SOD1G93A (H) EDL. At non-synaptic sites, GFP-positive cells do not express dsRed (hollow arrow; E, F, G, H) and dsRed-positive cells do not express GFP (filled arrow; E, F, G, H). Error bar = standard error of the mean. Scale bar = 10 μm (A–B), 50 μm (E–H), and 12.5 μm (E’-H’ and E’=H’).

We next assessed if S100β-GFP;NG2-dsRed mice could also be used to study PSCs at degenerating and regenerating NMJs. First, we examined expression of NG2-dsRed and S100β-GFP after crushing the fibular nerve (Dalkin et al., 2016). In this injury model, motor axons completely retract within 1 day and return to reinnervate vacated postsynaptic sites by 7 days post-injury in young adult mice. Similar to healthy uninjured EDL muscles (Figure 2E), NG2-dsRed and S100β-GFP were found co-expressed exclusively in PSCs at 4d and 7d post-injury (Figure 2F–G). Second, we crossed the SOD1G93A mouse line, (Gurney et al., 1994) a model of ALS shown to exhibit significant degeneration of NMJs (Moloney et al., 2014), with S100β-GFP;NG2-dsRed mice and examined the expression pattern of NG2-dsRed and S100β-GFP in the EDL during the symptomatic stage (P120). We again found NG2-dsRed and S100β-GFP co-expressed only in PSCs in the EDL of P120 SOD1G93A;S100β-GFP;NG2-dsRed mice (Figure 2H). Together, these data strongly indicate that this genetic labeling approach can be deployed to study the synaptic glia of the NMJ in a manner previously not possible in healthy and stressed NMJs.

To determine the relationship between NG2 expression and PSC differentiation, we analyzed NG2 expression in S100β-GFP+ Schwann cells during the course of NMJ development in the EDL muscle of S100β-GFP;NG2-dsRed mice (Figure 3A and Figure 3—figure supplement 1). We observed the presence of S100β-GFP+ cells at the NMJ as early as embryonic day 15 (E15) with 100% of NMJs having at least one S100β-GFP+ cell by post-natal day 9 (Figure 3A,B). During the embryonic developmental stages, we observed that NMJs are exclusively populated by S100β-GFP+ cells that do not express NG2-dsRed (Figure 3C). At post-natal day 0 (P0), however, we observed NG2-dsRed expression in a small subset of S100β-GFP+ cells (Figure 3A,C). Notably, the proportion of NMJs with S100β-GFP+;NG2-dsRed+ cells sharply increased between the ages of P0 and P9, coinciding with the period of NMJ maturation in mouse skeletal muscles (Figure 3C). By P21, when NMJ maturation in mice is near completion (Sanes and Lichtman, 1999), we observed S100β-GFP+;NG2-dsRed+ cells to be exclusively present at NMJs (Figure 3C). At this age, the number of S100β-GFP+;NG2-dsRed+ PSCs reached an average of 2.3 per NMJ and this remained unchanged in healthy young adult mice (Figure 3A,D). To confirm that dsRed expression from the NG2 promoter denotes the temporal and spatial transcriptional control of the NG2 gene in S100β-GFP;NG2-dsRed mice, we immunostained for NG2 protein. We found NG2 protein present at mature NMJs but not in NMJs of E18 mice with IHC (Figure 3—figure supplement 2). Thus, the induced expression of NG2 during the course of NMJ development in Schwann cells located proximally to the NMJ provides further evidence that NG2 is a marker of mature, differentiated S100β+ PSCs. It is possible that PSCs upregulate NG2 during development in order to restrict motor axon growth at the NMJ (Filous et al., 2014). Induced NG2 expression during NMJ development along with the constant presence of S100β-GFP+ cells (S100β-GFP+ or S100β-GFP+;NG2-dsRed+) and absence of single labeled NG2-dsRed+ cells at NMJs at every observed developmental time point (Figure 3B,C) strongly support previous studies indicating that PSCs originate from Schwann cells (Lee et al., 2017).

Figure 3 with 2 supplements see all
Analysis of NG2-dsRed distribution and PSC density during NMJ development in the EDL muscle.

(A) Representative images of NMJs, identified by nAChR clusters with fBTX (blue), in developing (E15, P0, P6, P9, P21) and adult S100β-GFP (green);NG2-dsRed (red) transgenic EDL. (B) The number of NMJs populated by at least one S100β-GFP+ cell increases between the ages of E15 and P9, at which point all observed NMJs have at least one S100β-GFP+ cell. (C) Analysis of NMJs that contain at least one single labeled S100β-GFP+ cell (green bar), at least one double labeled S100β-GFP+;NG2-dsRed+ cell (yellow bar) or a combination of single labeled S100β-GFP+ cells and double labeled S100β-GFP+;NG2-dsRed+ cells (green/yellow bar) shows that developing NMJs are exclusively populated by S100β-GFP+ cells in the embryonic stages and are increasingly populated by S100β-GFP+;NG2-dsRed+ cells as the NMJ develops. (D) The average number of PSCs per NMJ increases during development. Error bar = standard error of the mean. Scale bar = 10 μm. *=p < 0.05,***=P < 0.001; ****=P < 0.0001. Asterisks represent comparisons with P0 unless otherwise noted.

To gain insights into the rules that govern the distribution of PSCs at NMJs we compared PSC density in EDL, soleus, and diaphragm muscles to determine if PSC density is similar across muscles with varying NMJ sizes, fiber types and functional demands. Here, we observed similar PSC densities in each muscle type (Figure 4—figure supplement 1), suggesting that the number of PSCs directly correlates with the size of the NMJ and not the functional characteristics or fiber type composition of the muscles with which they are associated.

We next examined the spatial distribution of PSCs at the NMJ using the Nearest Neighbor (NN) analysis. This analysis measures the linear distance between neighboring cells in order to determine the regularity of spacing (Wassle and Riemann, 1978; Cook, 1996), quantified using the regularity index. In this analysis, randomly distributed groups of cells yield a nearest neighbor regularity index (NNRI) of 1.91 while those with nonrandom, regularly ordered distributions yield higher NNRI values (Reese and Keeley, 2015; Figure 4A). We found that the spacing of PSCs yielded high NNRI values and thus maintained ordered, non-random distributions at NMJs in the EDL muscle of adult mice. Moreover, this ordered distribution was maintained regardless of the overall number of PSCs at a given NMJ (Figure 4B–D). These observations are in accord with a published study indicating that PSCs occupy distinct territories at adult NMJs (Brill et al., 2011). In addition, these data strongly suggest that presynaptic, postsynaptic, and/or PSC-PSC mechanisms of communication dictate the spatial distribution of PSCs.

Figure 4 with 1 supplement see all
PSC distribution at the NMJ is non-random and ordered.

(A) The nearest neighbor (NN) distance, or the distance between a PSC and the closest neighboring PSC, is represented by the red line. The distance represented by the purple line is not considered for NN analysis. The distribution of NN values (shown in panels B, C, and D) is used to determine the degree of order in PSC distribution, as represented by the nearest neighbor regularity index (NNRI). Distribution patterns with an NNRI value greater than 1.91 are considered to be non-random. (B–D) Nearest neighbor distributions of S100-GFP+;NG2-dsRed+ PSCs in adult (P60) EDL muscle show that PSC distributions have orderly patterns of distribution with NNRI > 1.91 regardless of whether they are located at an NMJ with 2 PSCs (B), 3 PSCs (C), or 4 PSCs (D).

The ability to distinguish PSCs from all other Schwann cells makes it possible to identify genes that are either preferentially or specifically expressed in PSCs. We deployed fluorescence-activated cell sorting (FACS) to separately isolate double labeled S100β-GFP+;NG2-dsRed+ PSCs, single-labeled S100β-GFP+ Schwann cells, and single-labeled NG2-dsRed+ cells (including α-SMA pericytes and Tuj1+ precursor cells [Birbrair et al., 2013b]) from juvenile (P15-P22) S100β-GFP;NG2-dsRed transgenic mice. We then utilized RNA-Sequencing (RNA Seq) to compare the transcriptional profile of PSCs with the other two groups (Figure 5A). Light microscopy and expression analysis of GFP and dsRed using quantitative PCR (qPCR) confirmed that only cells of interest were sorted (Figure 5A–B). This analysis revealed a unique transcriptional profile for PSCs (Figure 5C). Notably, we found 567 genes enriched in PSCs that were not previously recognized to be associated with PSCs, glial cells or synapses (Supplementary file 1) using Ingenuity Pathway Analysis (IPA). We also found a number of genes preferentially expressed by PSCs with known roles at synapses (Mozer and Sandstrom, 2012; Fox and Umemori, 2006; Rafuse et al., 2000; Ranaivoson et al., 2019; Shapiro et al., 2007; Peng et al., 2010; Supplementary file 1). Providing additional insights about the function of PSCs, IPA revealed synaptogenesis, glutamate receptor, and axon guidance signaling as top canonical pathways under transcriptional regulation (Figure 5D).

Molecular analysis of S100β-GFP+;NG2-dsRed+ PSCs, S100β-GFP+ Schwan cells, and NG2-dsRed+ cells following isolation with FACS.

(A) Skeletal muscle from juvenile S100B-GFP;NG2-dsRed mice was dissociated and S100-GFP+;NG2-dsRed+ PSCs, S100β-GFP+ Schwan cells, and NG2-dsRed+ cells were sorted by FACS for RNA seq and qPCR. Representative fluorescence intensity gates for sorting of S100β-GFP+, NG2-dsRed+ and S100β-GFP+;NG2-dsRed+ cells are indicated in the scatter plot. GFP (y-axis) and dsRed (x-axis) fluorescence intensities were used to select gates for S100β-GFP+ cells (outlined in orange), NG2-dsRed+ cells (outlined in teal), and double labeled S100β-GFP+;NG2-dsRed+ cells (outlined in purple). Representative images of cells from sorted populations are shown. (B) GFP and dsRed qPCR was performed on FACS isolated cells to confirm specificity of sorting gates. (C) A heat map of RNA-seq results depicting genes with at least 5 counts and expression differences with a p-value of less than 0.01 between any 2 cell types reveals a distinct transcriptome in S100β-GFP+;NG2-dsRed+ PSCs versus S100β-GFP+ Schwann cells and NG2-dsRed+ cells. (D) Synaptogenesis and axon guidance signaling are among the most influential signaling pathways in PSCs according to Ingenuity Pathway Analysis of genes enriched in PSCs versus S100β-GFP+, and NG2-dsRed+ cells. (E) qPCR was performed on FACS isolated S100-GFP+;NG2-dsRed+ PSCs, S100β-GFP+ Schwan cells, and NG2-dsRed+ cells to verify mRNA levels of RNA seq identified PSC enriched genes. In each analysis, transcripts were not detected or detected at low levels in S100β-GFP+ Schwann cells and NG2-dsRed+ cells. Error bar = standard error of the mean. Scale bar = 10 µm.

We next cross-referenced our transcriptomic data with a list of genes compiled from published studies indicating enrichment or functional roles in PSCs (Young et al., 2005; Reynolds and Woolf, 1992; Robitaille, 1995; Robitaille et al., 1997; Rochon et al., 2001; Georgiou and Charlton, 1999; Trachtenberg and Thompson, 1996; Morris et al., 1999; Woldeyesus et al., 1999; Riethmacher et al., 1997; Personius et al., 2016; Park et al., 2017; Pinard et al., 2003; Descarries et al., 1998; Hess et al., 2007; Heredia et al., 2018; Darabid et al., 2018; Musarella et al., 2006; De Winter et al., 2006; Feng and Ko, 2008; Yang et al., 2001; Petrov et al., 2014; Robitaille et al., 1996; Gorlewicz et al., 2009; Wright et al., 2009). This analysis identified 27 genes expressed in isolated S100β-GFP+;NG2-dsRed+ PSCs that were previously shown to be associated with PSCs (Table 1). These included genes involved in detection and modulation of synaptic activity such as adenosine (Robitaille, 1995; Rochon et al., 2001), P2Y (Robitaille, 1995; Heredia et al., 2018; Darabid et al., 2018), acetylcholine (Robitaille et al., 1997; Petrov et al., 2014; Wright et al., 2009) and glutamate receptors (Pinard et al., 2003), Butyrylcholinesterase (BChE) (Petrov et al., 2014), and L-type calcium channels (Robitaille et al., 1996). Additionally, genes involved in NMJ development, synaptic pruning, and maintenance including agrin, 2',3'-cyclic nucleotide 3' phosphodiesterase (CNP) (Georgiou and Charlton, 1999), Erb-b2 receptor tyrosine kinase 2 (Erbb2) (Trachtenberg and Thompson, 1996; Morris et al., 1999; Woldeyesus et al., 1999), Erbb3 (Trachtenberg and Thompson, 1996; Riethmacher et al., 1997) GRB2-associated protein 1 (Gab1) (Park et al., 2017), myelin-associated glycoprotein (MAG) (Georgiou and Charlton, 1999), and myelin protein zero (Mpz) (Georgiou and Charlton, 1999) were detected in PSCs.

Table 1
Genes with functional roles in PSCs identified by RNA seq analysis of isolated PSCs.
GeneDescriptionProposed roleRead countLog2 change vs
NG2-dsRed+
Log2 change
vs S100β-GFP+
Reference
Adora2aAdenosine A2a receptorDetect/modulate synaptic activity8.1−3.68−2.67(Robitaille, 1995; Rochon et al., 2001)
Adora2bAdenosine A2b receptorDetect/modulate synaptic activity9.2−3.16−4.55(Robitaille, 1995; Rochon et al., 2001)
AgrnAgrinAChR aggregation2049.71.162.93(Georgiou and Charlton, 1999)
BcheButyrylcholinesteraseModulate synaptic ACh levels7191.07.897.21(Trachtenberg and Thompson, 1996)
Cacna1cL type Calcium channel, alpha 1 cDetect/modulate synaptic activity14.3−4.92−2.10(Morris et al., 1999)
Cacna1dL type Calcium channel, alpha 1dDetect/modulate synaptic activity18.4−0.42−1.49(Morris et al., 1999)
Cd44CD44 antigenMediates cell-cell interactions1249.20.75−1.22(Woldeyesus et al., 1999)
Chrm1Muscarinic acetylcholine receptor M1Detect/modulate synaptic activity14.8n.d.0.89(Robitaille et al., 1997; Riethmacher et al., 1997)
Cnp2',3'-cyclicnucleotide 3' phosphodiesteraseAnchors axon terminal at NMJ2990.24.231.66(Personius et al., 2016)
Erbb2Erb-b2 receptor tyrosine kinase 2Synaptogenesis/maintenance228.90.841.37(Park et al., 2017; Pinard et al., 2003; Descarries et al., 1998)
Erbb3Erb-b2 receptor tyrosine kinase 3Synaptogenesis/maintenance2471.37.054.46(Park et al., 2017; Hess et al., 2007)
GAb1GRB2-associated protein 1Synaptic pruning693.80.311.57(Heredia et al., 2018)
Grm1Glutamate receptor, metabotropic 1Detect/modulate synaptic activity9.2n.d.0.80(Darabid et al., 2018)
Grm5Glutamate receptor, metabotropic 5Detect/modulate synaptic activity38.0n.d.2.84(Darabid et al., 2018)
LNX1Ligand of numb-protein X 1Regulate NRG1 signaling37.5−2.29−0.70(Peper et al., 1974)
MAGMyelin-associated glycoproteinSynaptogenesis/maintenance136.03.12−0.55(Personius et al., 2016)
MpzMyelin protein zeroSynaptogenesis/maintenance4590.72.54−0.79(Personius et al., 2016)
Nos2Nitric oxide synthase 2, inducibleSynaptogenesis/modulate synaptic activity13.4−2.91−1.28(Musarella et al., 2006)
Nos3Nitric oxide synthase 3, endothelial cellSynaptogenesis/modulate synaptic activity48.6−2.69−0.68(Musarella et al., 2006)
P2ry1Purinergic receptor P2Y1Detect/modulate synaptic activity, synapse elimination144.40.522.21(Robitaille, 1995; De Winter et al., 2006; Feng and Ko, 2008)
P2ry2Purinergic receptor P2Y2Detect/modulate synaptic activity24.0−1.55−1.04(Robitaille, 1995)
P2ry10bP2Y receptor family member P2Y10bDetect/modulate synaptic activity10.0−1.25−3.14(Robitaille, 1995)
P2ry12P2Y receptor family member P2Y12Detect/modulate synaptic activity273.5n.d.3.70(Robitaille, 1995)
P2ry14P2Y receptor family member P2Y14Detect/modulate synaptic activity13.6−3.49−2.06(Robitaille, 1995)
S100bS100 protein betaIntracellular signaling1788.35.343.12(Reynolds and Woolf, 1992)
Sema3aSemaphorin 3aDetect/modulate synaptic activity136.62.951.07(Yang et al., 2001)
Tgfb1Transforming growth factor, beta 1AChR aggregation173.2−1.08−1.90(Petrov et al., 2014)

We deployed quantitative PCR (qPCR) to validate preferential expression of select genes in PSCs. To do so, we obtained RNA from S100β-GFP+;NG2-dsRed+ PSCs, single-labeled S100β-GFP+ Schwann cells, and single-labeled NG2-dsRed+ cells isolated using FACS from juvenile S100β-GFP;NG2-dsRed transgenic mice. We examined eight genes identified by RNA seq as being highly enriched in PSCs. This included newly identified genes (Ajap1, Col20a1, FoxD3, Nrxn1, PDGFa, and Pdlim4) and genes previously shown to be enriched (BChE [Petrov et al., 2014] and NCAM1 [Covault and Sanes, 1986]) in PSCs (Figure 5E). Validating RNA-Seq findings, qPCR analysis showed that all eight genes are highly enriched in PSCs compared to all other cell types isolated via FACS (Figure 5E). Additionally, immunostaining showed that NG2, a novel PSC-enriched gene identified by RNA-Seq, is concentrated at the NMJ (Figure 3—figure supplement 2).

Discussion

We have discovered a unique combination of molecular markers that allows us to specifically visualize, isolate, interrogate the transcriptome, and potentially alter the molecular composition of PSCs. We have shown that NG2 is specifically expressed by S100β-GFP+ PSCs but not myelinating S100β-GFP+ Schwann cells and thus the combined expression of S100β and NG2 is a unique molecular marker of PSCs in skeletal muscle. Providing further evidence that NG2 is a marker of differentiated PSCs, we have demonstrated that Schwann cells induce expression of NG2 shortly after they arrive at the NMJ during maturation of the synapse. However, the means by which the induced expression of NG2 is part of a program to establish and/or further specify PSC identity in Schwann cells at the NMJ, through activation of the NG2 promoter, remains to be determined.

We utilized FACS to isolate S100β-GFP+;NG2-dsRed+ PSCs from skeletal muscle to analyze the PSC transcriptome. To our knowledge this is the first time that the transcriptome of PSCs, or any other type of glial cell that associates exclusively with synapses, has been characterized. This analysis reveals expression of a number of genes that have been previously implicated in modulation of synaptic activity, synaptic pruning, and synaptic maintenance by PSCs. We identified a number of novel genes that are highly expressed in PSCs but not Schwann cells or NG2+ cells. These genes have the potential to assist PSC research by serving as molecular markers that can be utilized for PSC-specific genetic manipulations, PSC ablation, and isolation of PSCs for cell culture and molecular analysis. We verified a number of these with qPCR and IHC. This analysis, therefore, reveals a unique gene expression signature that distinguishes PSCs from all other Schwann cells.

While the role of the majority of genes found enriched in PSCs at the neuromuscular synapse remains to be determined, it is worth noting that many have been shown to play key roles in neuronal circuits in the central nervous system and in cell-cell communication. This is the case for NG2 which has been shown to terminate axonal growth in glial scars in the spinal cord (Filous et al., 2014). Therefore, it is possible that NG2 is utilized by PSCs to tile, and thus occupy unique territories, and prevent motor axons from developing sprouts that extend beyond the postsynaptic partner. Supporting this possibility, we have found that the NG2 promoter is active in some PSCs at P0 (Figure 3C), a time when motor axon nerve endings at NMJs undergo rapid morphological changes (Sanes and Lichtman, 1999; Sanes and Lichtman, 2001). The progressive activation of the NG2 promoter in PSCs is complete by P9 (Figure 3C), which coincides with the elimination of extranumeral axons innervating the same postsynaptic site in mice (Sanes and Lichtman, 1999; Sanes and Lichtman, 2001). Therefore, PSCs may utilize NG2 to promote the maturation of the presynaptic region and thus the NMJ. Furthermore, PSCs may utilize NG2 to repel each other as they tile during development to occupy unique territories at the NMJ (Brill et al., 2011).

With these tools, it is now possible to determine which cellular and molecular determinants are critical for PSC differentiation, maturation, and function at the NMJ. It will also allow us to ascertain the contribution of PSCs to NMJ repair following injury and NMJ degeneration during normal aging and the progression of neuromuscular diseases, such as Amyotrophic Lateral Sclerosis (ALS) and Spinal Muscular Atrophy (SMA). Our strategy of specifically labeling synaptic glia, using a combination of protein markers uniquely expressed in this cell type, may serve as a springboard for unprecedented approaches for studying not only PSC function at the NMJ, but also synapse-associated glia throughout the CNS. Indeed, we have observed subsets of astrocytes in the brain that co-express both S100β and NG2, as has been previously reported in the context of a lineage tracing analysis (Deloulme et al., 2004). Future studies will determine the generality of our approach in discerning the functional roles of synaptic glia in the development, maintenance, and function of select synapses.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Genetic reagent (M. musculus)S100b-GFPPMID:15590915MGI:3588512Dr. Wesley Thompson (Texas A and M)
Genetic reagent (M. musculus)NG2-dsRedPMID:18045844MGI:3796063Dr. Akiko Nishiyama (University of Connecticut)
Genetic reagent (M. musculus)SOD1G93APMID:8209258MGI:2183719Dr. Deng (Northwestern
University)
AntibodyGuinea pig polyclonal anti-NG2PMID:19058188Antibody Registry: AB_25722991:250
AntibodyAlexa Fluor-488 goat polycolonal anti guinea pigInvitrogenRRID:AB_25341171:1000
AntibodyAlexa Fluor-488 goat polyclonal anti rabbitInvitrogenCatalog# A-110081:1000
Software, algorithmIngenuity Pathway AnalysisQiagenRRID:SCR_008117
Software, algorithmGraphPad PrismGraphPadRRID:SCR_002798
Software, algorithmRThe R Project for Statistical ComputingRRID:SCR_001905
Software, algorithmImageJImageJRRID:SCR_003070
Software, algorithmBio-Rad CFX ManagerBio-RadRRID:SCR_017251
Commercial assay or kitPicoPure RNA Isolation KitThermoFisherCatalog#KIT0204
Commercial assay or kitiScript cDNA synthesis kitBio-RadCatalog#1708891
Commercial assay or kitSsoAdvanced PreAmp SupermixBio-RadCataolog#1725160
Commercial assay or kitiTAQ Univeral SYBR Green SupermixBio-RadCatalog#1725121
Chemical compound, drugAlexa Fluor-555 alpha- bungarotoxinInvitrogenCatalog#B35451
Chemical compound, drugDAPIThermoFisherCatalog#D1306

Mice

SOD1G93A98 (Gurney et al., 1994), S100β-GFP (B6;D2-Tg(S100b-EGFP)1Wjt/J) (Zuo et al., 2004) and NG2-dsRed mice (Tg(Cspg4-DsRed.T1)1Akik/J) (Zhu et al., 2008) were obtained from Jackson Labs (Bar Harbor, ME). S100β-GFP and NG2-dsRed mice were crossed to generate S100β-GFP;NG2-dsRed mice. Offspring were genotyped using Zeiss LSM900 to check for fluorescent labels. SOD1G93A mice were crossed with S100β-GFP;NG2-dsRed mice to generate S100β-GFP;NG2-dsRed;SOD1G93A mice. Postnatal mice older than 9 days of age were anesthetized and immediately perfused with 4% paraformaldehyde (PFA) overnight. Pups were anesthetized by isoflurane and euthanized by cervical dislocation prior to muscle dissociation. Adult mice were anesthetized using CO2 and then perfused transcardially with 10 ml of 0.1 m PBS, followed by 25 ml of ice‐cold 4% PFA in 0.1 m PBS (pH 7.4). All experiments were carried out under NIH guidelines and animal protocols approved by the Brown University and Virginia Tech Institutional Animal Care and Use Committee.

Fibular nerve crush

Request a detailed protocol

Adult S100β-GFP;NG2-dsRed mice were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg) delivered intraperitoneally. The fibular nerve was crushed at its intersection with the lateral tendon of the gastrocnemius muscle using fine forceps, as described previously (Dalkin et al., 2016). Mice were monitored for 2 hr following surgery and administered buprenorphine (0.05-.010 mg/kg) at 12 hr intervals during recovery.

Immunohistochemistry and NMJ visualization

Request a detailed protocol

For NG2 immunohistochemistry (IHC), muscles were incubated in blocking buffer (5% lamb serum, 3% BSA, 0.5% Triton X-100 in PBS) at room temperature for 2 hr, incubated with anti-NG2 antibody (courtesy of Dr. Dwight Bergles) diluted at 1:250 in blocking buffer overnight at 4°C, washed 3 times with 0.1M PBS for 5 min. Muscles were then incubated with 1:1000 Alexa Fluor-488 conjugated anti-rabbit or guinea pig antibody (A-11008, Invitrogen, Carlsbad, CA) and 1:1000 Alexa Fluor-555 conjugated α-bungarotoxin (fBTX; Invitrogen, B35451) in blocking buffer for 2 hr at room temperature and washed 3 times with 0.1M PBS for 5 min. For all other NMJ visualization, muscles were incubated in Alexa Fluor-647 conjugated α-bungarotoxin (fBTX; Invitrogen, B35450) at 1:1000 and 4',6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI; D1306, ThermoFisher, Waltham, MA) at 1:1000 in 0.1M PBS at 4°C overnight. Muscles were then washed with 0.1M PBS 3 times for 5 min each. Muscles were whole mounted using Vectashield (H-1000, Vector Labs, Burlingame, CA) and 24 × 50–10.5 cover glass (ThermoFisher).

Confocal microscopy of PSCs and NMJs

Request a detailed protocol

All images were taken with a Zeiss LSM700, Zeiss LSM 710, or Zeiss LSM 900 confocal light microscope (Carl Zeiss, Jena, Germany) with a 20 × air objective (0.8 numerical aperture), 40 × oil immersion objective (1.3 numerical aperture), or 63 × oil immersion objective (1.4 numerical aperture) using the Zeiss Zen Black software. Optical slices within the z‐stack were taken at 1.00 μm or 2.00 μm intervals. High resolution images were acquired using the Zeiss LSM 900 with Airyscan under the 63 × oil immersion objective in super resolution mode. Optical slices within the z-stack were 0.13 μm with a frame size of 2210 × 2210 pixels. Images were collapsed into a two‐dimensional maximum intensity projection for analysis.

Image analysis

NMJ size

Request a detailed protocol

To quantify the area of NMJs, the area of the region occupied by nicotinic acetylcholine receptors (nAChRs, labeled by fBTX, was measured using ImageJ software. At least 100 nAChRs were analyzed for number of fragments, individual nicotinic acetylcholine receptor (nAChR) clusters, from each muscle to represent an individual mouse. At least 3 animals per age group were analyzed to generate the represented data.

Cells associated with NMJs

Request a detailed protocol

Cell bodies were visualized via GFP and/or dsRed signal, and were confirmed as cell bodies by the presence of a DAPI+ nucleus. The area of each cell body was measured by tracing the outline of the entire cell body using the freehand tool in ImageJ. To quantify the number of cells associated with NMJs, the number of cell bodies directly adjacent to each NMJ was counted. Every cell that overlapped with or directly abutted the fBTX signal was considered adjacent to the NMJ. At least 3 animals per age group were analyzed to generate the represented data. Cells were examined in at least 100 NMJs from each muscle to represent an individual mouse.

Spacing of PSCs at NMJs

Request a detailed protocol

NMJs were identified via fBTX signal. PSCs were identified by the colocalization of GFP, dsRed, and DAPI signal in addition to their location at NMJs. The area of each PSC and NMJ was measured. The linear distance from the center of each PSC soma to the center of the nearest PSC soma at a single NMJ was measured. The distances were then separated into 5 μm bins and plotted in a histogram. All linear measurements were made using the line tool in the ImageJ software. At least 100 NMJs were analyzed from each muscle to represent an individual mouse.

Muscle dissociation and fluorescence activated cell sorting

Request a detailed protocol

Diaphragm, pectoralis, forelimb and hindlimb muscles were collected from P15-P21 S100β-GFP;NG2-dsRed mice. After removal of connective tissue and fat, muscles were cut into 5 mm2 pieces with forceps and digested in 2 mg/mL collagenase II (Worthington Chemicals, Lakewood, NJ) for 1 hr at 37°C. Muscles were further dissociated by mechanical trituration in Dulbecco’s modified eagle medium (Life Technologies, Carlsbad, CA) containing 10% horse serum (Life Technologies) and passed through a 40 µm filter to generate a single cell suspension. Excess debris was removed from the suspension by centrifugation in 4% BSA followed by a second centrifugation in 40% Optiprep solution (Sigma-Aldrich, St. Louis, MO) from which the interphase was collected. Cells were diluted in FACS buffer containing 1 mM EDTA, 25 mM Hepes, 1% heat inactivated fetal bovine serum (Life Technologies), in Ca2+/Mg2+ free 1X Dulbecco’s phosphate buffered saline (Life Technologies). FACS was performed with a Sony SH800 Cell Sorter (Sony Biotechnology, San Jose, CA). Representative fluorescence intensity gates for sorting of S100-GFP+, NG2-dsRed+ and S100-GFP+;NG2-dsRed+ cells are provided in Figure 5. Purity of the sorted cell population was confirmed by visual inspection of sorted cells using an epifluorescence microscope and with dsRed and GFP qPCR (Figure 5A,B). A single mouse was used for each replicate and an average of 7500 cells per replicate were collected for each cell group.

RNA-seq and qPCR

Request a detailed protocol

RNA was isolated from S100β-GFP+, NG2-dsRed+, or S100β-GFP+/NG2-dsRed+ cells following FACS with the PicoPure RNA Isolation Kit (ThermoFisher). The maximum number of cells that could be collected by FACS following dissociation of muscles collected from one mouse was used as a single replicate. On average, a single replicate consisted of 7,500 cells. RNA seq was performed by Genewiz on 12 replicates per cell type. Following sequencing, data were trimmed for both adaptor and quality using a combination of ea-utils and Btrim (Aronesty, 2013; Kong, 2011). Sequencing reads were aligned to the genome using HiSat2 (Kim et al., 2019) and counted via HTSeq (Anders et al., 2015). QC summary statistics were examined to identify any problematic samples (e.g. total read counts, quality and base composition profiles (+/- trimming), raw fastq formatted data files, aligned files (bam and text file containing sample alignment statistics), and count files (HTSeq text files). Following successful alignment, mRNA differential expression were determined using contrasts of and tested for significance using the Benjamini-Hochberg corrected Wald Test in the R-package DESeq2 (Love et al., 2014). Failed samples were identified by visual inspection of pairs plots and removed from further analysis resulting in the following number of replicates for each cell type: NG2-dsRed+, 10; S100β-GFP+, 7; NG2-dsRed+;S100β-GFP+, 9. Functional and pathway analysis was performed using Ingenuity Pathway Analysis (QIAGEN Inc, https://www.qiagenbio-informatics.com/products/ingenuity-pathway-analysis). Confirmation of expression of genes identified by RNA-seq was performed on 6 additional replicates of each cell type using quantitative reverse transcriptase PCR (qPCR). Reverse transcription was performed with iScript (Bio-Rad, Hercules, CA) and was followed by a preamplification PCR step with SsoAdvanced PreAmp Supermix (Bio-Rad) prior to qPCR using iTAQ SYBR Green and a CFX Connect Real Time PCR System (Bio-Rad). Relative expression was normalized to 18S using the 2-ΔΔCT method. The primers used for both preamplification and qPCR are listed in Supplementary file 2.

Statistics

An unpaired t-test or one-way ANOVA with Bonferroni post hoc analysis were used for statistical evaluation. Data are expressed as the mean ± standard error of the mean, and p<0.05 was considered statistically significant. The number of replicates is as follows: RNA seq, 7–10; qPCR, 6; comparisons between EDL, soleus, and diaphragm, 1; comparison between SOD1G93A and healthy adult muscle, 1; all other analyses, 3. Statistical analyses were performed using GraphPad Prism8 and R. Data values and p-values are reported within the text.

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
  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
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
    Characterization and localization of the mr = 43,000 proteins associated with acetylcholine receptor-rich membranes
    1. S Porter
    2. SC Froehner
    (1983)
    The Journal of Biological Chemistry 258:10034–10040.
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
  81. 81
  82. 82
  83. 83
  84. 84
  85. 85
  86. 86
  87. 87
    Structural and functional properties of the acetylcholine regulator
    1. A Sobel
    2. J Hofler
    3. T Heidmann
    4. JP Changeux
    (1979)
    Advances in Cytopharmacology 3:191–196.
  88. 88
  89. 89
  90. 90
  91. 91
  92. 92
  93. 93
  94. 94
  95. 95
  96. 96
  97. 97
  98. 98
  99. 99

Decision letter

  1. Beth Stevens
    Reviewing Editor; Boston Children's Hospital, United States
  2. Gary L Westbrook
    Senior Editor; Oregon Health and Science University, United States
  3. Chien-Ping Ko
    Reviewer

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

Acceptance summary:

Perisynaptic Schwann cells (PSCs) at the neuromuscular junction (NMJ) play essential roles in synaptic function, formation, maintenance, and repair. However, the molecular characteristics specific to PSCs are largely unknown. In this manuscript, Castro et al., have adopted a novel approach of combining S100B and NG2 labels to reveal a specific probe for PSCs and identify several associated novel genes. Their new discoveries advance molecular understanding of synapse-glia interactions at the NMJ in health and disease, and provide a foundation for further exploration.

Decision letter after peer review:

Thank you for submitting your article "Specific labeling reveals unique cellular and molecular features of synaptic Schwann cells" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Gary Westbrook as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Chien-Ping Ko (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.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data. Our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper.

Summary:

The manuscript by Castro et al. describes the use of combined S100b and NG2 to identify perisynaptic (terminal) Schwann cells at the neuromuscular junction. The role of PSCs in the development, function, degeneration and regeneration of the NMJ is well established, but their careful study has been limited by a lack of molecular markers. This study convincingly demonstrates that combined expression of S100b (labeling all Schwann cells) and NG2, labeling only relatively mature terminal Schwann cells (as well as other cell types in muscle) can be combined to study, and more importantly isolate, PSCs. Some data presented largely reproduce previous findings regarding PSCs, but this information serves to validate the use of S100b/NG2 to identify this population. Given the long-standing interest in identifying specific markers of PSCs to allow their isolation and manipulation, these findings are important for the field. However, a few items would improve the manuscript and data presentation. as guided by the Essential revisions section below.

Essential revisions:

1) Address each of the key comments (below) in the revised paper, especially regarding the potential limitations and uncertainties of the new marker, as raised by both reviewers.

2) Transcriptional Profiling of PSCs

– More could be done to validate the top DE genes using RNA scope and in situ hybridization. That is relatively straightforward technically if this is possible during covid19 period. For example, the authors provide only one IHC for validation beyond the qPCR and it is not clear why an adherens junction (Ajap1) would show diffuse cytoplasmic localization, so more will need to be done to validate this antibody and show that immunoreactivity is indeed AJAP1 (in addition to validating other DE genes). The authors are referred to the above paragraph regarding eLife policy during COVID-19 for our expectations.

– Are there other NG2+ cells in muscle and if so, this needs to be considered in the interpretation of the current RNA-seq data from this population?

– The presentation of the RNA-seq data could be formatted to be a little more useful to help rank candidates of interest, including read count and fold change (similar format to Table 1). See comments reviewer 2.

Revisions expected in follow-up work:

– Additional validation and spatial mapping of more DE expressed genes (sm FISH with combinations of markers).

Reviewer #1:

Comments have been incorporated into the editorial comments above.

Reviewer #2:

Perisynaptic Schwann cells (PSCs) at the neuromuscular junction (NMJ) play essential roles in synaptic function, formation, maintenance, and repair. However, the molecular characteristics specific to PSCs are largely unknown. In this manuscript, Castro et al. have adopted a novel approach of combining S100B and NG2 labels to reveal a specific probe for PSCs and identify several associated novel genes. The manuscript is well-written and comprehensive. Their findings are truly exciting, compelling, and immensely significant, as the field desperately needs specific probes for PSCs. Their new discoveries would advance molecular understanding of synapse-glia interactions at the NMJ in health and disease.

1) Introduction, last paragraph and Results, first paragraph. The authors may like to comment on whether they are the first group to show NG2 at PSCs and/or how they came up with the idea of combining both S100B and NG2 for their approach to identify PSCs. It would also be useful to comment/speculate the potential function of NG2 and other genes such as Ajap1 in PSCs.

2) Results, first paragraph, Figure 1F. The authors use the AChR fragmentation index to assess if NMJ deterioration occurs in their transgenic mouse line. It would be more sensitive and informative by also labeling the presynaptic nerve terminals to check if there is denervation. In addition, it is desirable to examine the quantal contents to confirm if indeed there are no functional deficits at the NMJ in their transgenic mice.

3) It is unclear about the statement, "we failed to find.… throughout the course of embryonic and postnatal development." But NG2 is expressed at least partially starting from P0 (Figure 2C). The authors may also like to comment/speculate on the reason/significance for NG2 expression only for mature differentiated PSCs.

4) The authors used FACS to isolate S100B-GFP+;NG2-dsRed+ PSCs from skeletal muscles to analyze the PSC transcriptome. However, it appears that NG2 expresses only in PSCs after birth and in mature PSCs (Figure 2C, Figure 2—figure supplement 1). It is also not known whether denervation after nerve injury and subsequent repair would alter the expression of NG2. If NG2 is downregulated or absent in denervated PSCs, their new probes for PSCs and associated genes might not be able to address the role of PSCs after nerve injury. It would be desirable to perform a relatively straightforward experiment by simply examining NG2 staining after nerve injury. This additional experiment would broaden the applications of using their transgenic line to address a key role of PSCs in synaptic repair in future studies. Otherwise, their approaches would be somewhat limited to studying only the mature PSCs in intact muscles.

Reviewer #3:

The paper by Castro et al. describes the use of combined S100b and NG2 to identify perisynaptic (terminal) Schwann cells at the neuromuscular junction. Besides the identification of this marker combination, the most significant results presented are the transcriptional profile of these cells isolated by flow cytometry from muscle. Given the long-standing interest in identifying specific markers of PSCs to allow their isolation and manipulation, these findings are important for the field. The image quality in the paper is excellent, and the quantification and statistics are rigorous and convincing. A few items would further improve the manuscript and data presentation.

The authors should comment on what the other NG2+ cells in muscle may be (recognizing there may be multiple cell types). This would help in future studies attempting to use a similar intersectional approach to study PSCs and also in the interpretation of the current RNA-seq data from this population (indeed, the -seq may help with the identification of these cell types).

The list of genes in Supplementary file 1 that also show enrichment in the S100b+/NG2+ cells versus either single marker is a potentially very useful resource for the field. As such, it should include some additional information to help rank candidates of interest, including read count and fold change (similar format to Table 1, without references and proposed functions).

There is a lot of future potential to be based on this work (response to denervation/axotomy, status and gene expression changes in disease, genetic ablation, etc.). Addition of such studies would strengthen this paper, but also require significant time and effort and exactly which of these studies would be most important and most informative is hard to predict. Therefore, the intrinsic value of publishing the identification of these markers and the list of PSC-enriched genes is noted as it will allow the field to subsequently pursue these additional experiments.

https://doi.org/10.7554/eLife.56935.sa1

Author response

Essential revisions:

1) Address each of the key comments (below) in the revised paper, especially regarding the potential limitations and uncertainties of the new marker, as raised by both reviewers.

2) Transcriptional Profiling of PSCs

– More could be done to validate the top DE genes using RNA scope and in situ hybridization. That is relatively straightforward technically if this is possible during covid19 period. For example, the authors provide only one IHC for validation beyond the qPCR and it is not clear why an adherens junction (Ajap1) would show diffuse cytoplasmic localization, so more will need to be done to validate this antibody and show that immunoreactivity is indeed AJAP1 (in addition to validating other DE genes). The authors are referred to the above paragraph regarding eLife policy during COVID-19 for our expectations.

We thank the reviewer for this comment, and agree that further validation of differentially expressed genes would strengthen this dataset. However, our lab does not currently possess the proper tools and equipment to perform in situ hybridization and/or RNA scope. Obtaining and setting up this equipment would take considerable time, and would require the purchase of new materials and equipment. It is unlikely that this will be possible in the immediate future due to COVID-19 restrictions, which for example continues to keeps us away from our lab and will make it difficult to purchase additional items not deemed necessary. Please do note that we do intend to obtain the equipment necessary to begin performing these and other important techniques for assessing PSCs in our lab once we resume “normal” research activities and as funds become available. However, we hope that our current validation using immunohistochemistry and qPCR will be sufficient at this time. Given that it will not be possible to validate our Ajap1 antibody with Ajap1 knockout tissue, we have removed the Ajap1 IHC results from the manuscript (changes made to Figure 5 and Results, last paragraph).

– Are there other NG2+ cells in muscle and if so, this needs to be considered in the interpretation of the current RNA-seq data from this population?

The reviewer has brought attention to an area of the manuscript that is currently lacking in sufficient detail. We have updated the text to include a better account of the types of cells that express NG2 in muscles based on published work by Birbrair A, et al., 2013a, b (changes made to Results, first and eighth paragraphs). We have also updated the text to reflect the fact that the collection of NG2-dsRed+ cells from muscles for RNA-seq represents a heterogeneous population of cell-types.

– The presentation of the RNA-seq data could be formatted to be a little more useful to help rank candidates of interest, including read count and fold change (similar format to Table 1). See comments reviewer 2.

We have modified the presentation of our RNA-seq data in Supplementary file 1 to include read count and fold change. Our entire RNA-seq data-set has been deposited in the NCBI-GEO database as well, so all of this information will be readily available to the reader.

Revisions expected in follow-up work:

– Additional validation and spatial mapping of more DE expressed genes (sm FISH with combinations of markers).

Reviewer #1:

Comments have been incorporated into the editorial comments above.

Reviewer #2:

[…] 1) Introduction, last paragraph and Results, first paragraph. The authors may like to comment on whether they are the first group to show NG2 at PSCs and/or how they came up with the idea of combining both S100B and NG2 for their approach to identify PSCs. It would also be useful to comment/speculate the potential function of NG2 and other genes such as Ajap1 in PSCs.

We thank the reviewer for pointing out this important omission. This manuscript concerns the discovery of new molecular markers, and we failed to explain in sufficient detail how that discovery was made. We have updated the text to rectify this mistake (changes made to Results, first paragraph). We have also expanded the text to include a more detailed discussion of the possible roles of NG2 in NMJ development (changes made to Discussion, third paragraph).

2) Results, first paragraph, Figure 1F. The authors use the AChR fragmentation index to assess if NMJ deterioration occurs in their transgenic mouse line. It would be more sensitive and informative by also labeling the presynaptic nerve terminals to check if there is denervation. In addition, it is desirable to examine the quantal contents to confirm if indeed there are no functional deficits at the NMJ in their transgenic mice.

We agree with the reviewer that analysis of presynaptic nerve terminals and quantal content would contribute to our analysis here, and we cannot discount the possibility that there are currently unidentified negative consequences of the co-expression of GFP and dsRed in PSCs. Unfortunately, performing these additional analyses would take a great deal of time and resources, and would be difficult to complete in a timely manner as long as COVID-19 restrictions persist. In this revision, we have added that we see no discernible differences between general features of S100B-GFP+ animals and NG2dsRed+/S100B-GFP+ animals, such as overall health or basic motor capacity (changes made to Results, second paragraph). We have also added a sentence that discusses this particular limitation of the study. Additionally, we see no changes in PSC number (Figure 1E) between NG2-dsRed+ and NG2-dsRed- animals.

3) It is unclear about the statement, "we failed to find.… throughout the course of embryonic and postnatal development." But NG2 is expressed at least partially starting from P0 (Figure 2C). The authors may also like to comment/speculate on the reason/significance for NG2 expression only for mature differentiated PSCs.

We agree with the reviewer that we were unclear in the quoted section of the text. We have updated the text to clarify when NG2 appears in PSCs. Additionally, we included text describing the possible function of NG2 at the NMJ (changes made to Results, fourth paragraph).

4) The authors used FACS to isolate S100B-GFP+;NG2-dsRed+ PSCs from skeletal muscles to analyze the PSC transcriptome. However, it appears that NG2 expresses only in PSCs after birth and in mature PSCs (Figure 2C, Figure 2—figure supplement 1). It is also not known whether denervation after nerve injury and subsequent repair would alter the expression of NG2. If NG2 is downregulated or absent in denervated PSCs, their new probes for PSCs and associated genes might not be able to address the role of PSCs after nerve injury. It would be desirable to perform a relatively straightforward experiment by simply examining NG2 staining after nerve injury. This additional experiment would broaden the applications of using their transgenic line to address a key role of PSCs in synaptic repair in future studies. Otherwise, their approaches would be somewhat limited to studying only the mature PSCs in intact muscles.

We agree with the reviewer and have added additional data and information that demonstrate the wide-spread utility of these markers to study PSCs at NMJs under various stressors (Figure 2E-H). This includes data showing that PSCs continue to be the only cells expressing both GFP and dsRed (proxies for S100B and NG2) after crushing innervating motor axons and during the symptomatic stage of ALS in a mouse model of the disease characterized by significant destruction of NMJs (changes made to Results, third paragraph). These data strongly indicate that these markers will be sufficient to study PSCs in diseases, injury and potentially during aging.

Reviewer #3:

The paper by Castro et al. describes the use of combined S100b and NG2 to identify perisynaptic (terminal) Schwann cells at the neuromuscular junction. Besides the identification of this marker combination, the most significant results presented are the transcriptional profile of these cells isolated by flow cytometry from muscle. Given the long-standing interest in identifying specific markers of PSCs to allow their isolation and manipulation, these findings are important for the field. The image quality in the paper is excellent, and the quantification and statistics are rigorous and convincing. A few items would further improve the manuscript and data presentation.

The authors should comment on what the other NG2+ cells in muscle may be (recognizing there may be multiple cell types). This would help in future studies attempting to use a similar intersectional approach to study PSCs and also in the interpretation of the current RNA-seq data from this population (indeed, the RNA-seq may help with the identification of these cell types).

We agree, see response to Essential revisions.

The list of genes in Supplementary file 1 that also show enrichment in the S100b+/NG2+ cells versus either single marker is a potentially very useful resource for the field. As such, it should include some additional information to help rank candidates of interest, including read count and fold change (similar format to Table 1, without references and proposed functions).

We agree, see response to Essential revisions.

There is a lot of future potential to be based on this work (response to denervation/axotomy, status and gene expression changes in disease, genetic ablation, etc.). Addition of such studies would strengthen this paper, but also require significant time and effort and exactly which of these studies would be most important and most informative is hard to predict. Therefore, the intrinsic value of publishing the identification of these markers and the list of PSC-enriched genes is noted as it will allow the field to subsequently pursue these additional experiments.

https://doi.org/10.7554/eLife.56935.sa2

Article and author information

Author details

  1. Ryan Castro

    1. Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, Providence, United States
    2. Center for Translational Neuroscience, Robert J. and Nancy D. Carney Institute for Brain Science and Brown Institute for Translational Science, Brown University, Providence, United States
    3. Neuroscience Graduate Program, Brown University, Providence, United States
    Contribution
    Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2316-8039
  2. Thomas Taetzsch

    1. Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, Providence, United States
    2. Center for Translational Neuroscience, Robert J. and Nancy D. Carney Institute for Brain Science and Brown Institute for Translational Science, Brown University, Providence, United States
    Contribution
    Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-3257-1142
  3. Sydney K Vaughan

    1. Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, Providence, United States
    2. Center for Translational Neuroscience, Robert J. and Nancy D. Carney Institute for Brain Science and Brown Institute for Translational Science, Brown University, Providence, United States
    Contribution
    Data curation, Formal analysis, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-3427-4654
  4. Kerilyn Godbe

    Fralin Biomedical Research Institute at Virginia Tech Carilion, Roanoke, United States
    Contribution
    Formal analysis, Writing - review and editing
    Competing interests
    No competing interests declared
  5. John Chappell

    Fralin Biomedical Research Institute at Virginia Tech Carilion, Roanoke, United States
    Contribution
    Resources, Writing - review and editing
    Competing interests
    No competing interests declared
  6. Robert E Settlage

    Department of Advanced Research Computing, Virginia Tech, Blacksburg, United States
    Contribution
    Formal analysis, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1354-7609
  7. Gregorio Valdez

    1. Department of Molecular Biology, Cellular Biology, and Biochemistry, Brown University, Providence, United States
    2. Center for Translational Neuroscience, Robert J. and Nancy D. Carney Institute for Brain Science and Brown Institute for Translational Science, Brown University, Providence, United States
    3. Department of Neurology, Warren Alpert Medical School of Brown University, Providence, United States
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Writing - review and editing
    For correspondence
    gregorio_valdez@brown.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0375-4532

Funding

National Institutes of Health (R01AG055545)

  • Gregorio Valdez

National Institutes of Health (R56AG051501)

  • Gregorio Valdez

National Institutes of Health (R21NS106313)

  • Gregorio Valdez

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

Acknowledgements

We thank members of the Valdez laboratory for providing helpful comments throughout the course of this project.

Ethics

Animal experimentation: All experiments were carried out under NIH guidelines and animal protocols approved by the Brown University (IACUC# 19-05-0013) and Virginia Tech (IACUC# 18-148 and 18-176) Institutional Animal Care and Use Committee.

Senior Editor

  1. Gary L Westbrook, Oregon Health and Science University, United States

Reviewing Editor

  1. Beth Stevens, Boston Children's Hospital, United States

Reviewer

  1. Chien-Ping Ko

Publication history

  1. Received: April 8, 2020
  2. Accepted: June 8, 2020
  3. Version of Record published: June 25, 2020 (version 1)
  4. Version of Record updated: July 2, 2020 (version 2)

Copyright

© 2020, Castro 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

  • 1,041
    Page views
  • 142
    Downloads
  • 1
    Citations

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

Further reading

    1. Biochemistry and Chemical Biology
    2. Cell Biology
    Mark Shaaya et al.
    Tools and Resources Updated

    Engineered allosteric regulation of protein activity provides significant advantages for the development of robust and broadly applicable tools. However, the application of allosteric switches in optogenetics has been scarce and suffers from critical limitations. Here, we report an optogenetic approach that utilizes an engineered Light-Regulated (LightR) allosteric switch module to achieve tight spatiotemporal control of enzymatic activity. Using the tyrosine kinase Src as a model, we demonstrate efficient regulation of the kinase and identify temporally distinct signaling responses ranging from seconds to minutes. LightR-Src off-kinetics can be tuned by modulating the LightR photoconversion cycle. A fast cycling variant enables the stimulation of transient pulses and local regulation of activity in a selected region of a cell. The design of the LightR module ensures broad applicability of the tool, as we demonstrate by achieving light-mediated regulation of Abl and bRaf kinases as well as Cre recombinase.

    1. Cell Biology
    2. Microbiology and Infectious Disease
    Harvie P Portugaliza et al.
    Research Article

    Malaria transmission is dependent on the formation of gametocytes in the human blood. The sexual conversion rate, the proportion of asexual parasites that convert into gametocytes at each multiplication cycle, is variable and reflects the relative parasite investment between transmission and maintaining the infection. The impact of environmental factors such as drugs on sexual conversion rates is not well understood. We developed a robust assay using gametocyte-reporter parasite lines to accurately measure the impact of drugs on sexual conversion rates, independently from their gametocytocidal activity. We found that exposure to subcurative doses of the frontline antimalarial drug dihydroartemisinin (DHA) at the trophozoite stage resulted in a ~ fourfold increase in sexual conversion. In contrast, no increase was observed when ring stages were exposed or in cultures in which sexual conversion was stimulated by choline depletion. Our results reveal a complex relationship between antimalarial drugs and sexual conversion, with potential public health implications.