Skip to main content
PMC home page
eLife logoLink to eLife
. 2020 Jun 25;9:e56935. doi: 10.7554/eLife.56935

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

Ryan Castro 1,2,3, Thomas Taetzsch 1,2, Sydney K Vaughan 1,2, Kerilyn Godbe 4, John Chappell 4, Robert E Settlage 5, Gregorio Valdez 1,2,6,✉
Editors: Beth Stevens7, Gary L Westbrook8
PMCID: PMC7316509  PMID: 32584256

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.

Research organism: Mouse

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.

Figure 1. Co-expression of S100β and NG2 is unique to PSCs in the EDL muscle.

Figure 1.

Open in a new tab

(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.

Figure 2. The NG2-DsRed/S100β-GFP mouse line can be used to reliably identify PSCs in healthy and stressed NMJs.

Figure 2.

Open in a new tab

(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. 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.

Figure 3.

Figure 3—figure supplement 1. Color and grayscale images of PSCs in the EDL muscle of (A) E15, (B) E18, (C) P0, (D) P6, (E) P9, (F) P21, and (G) adult S100β-GFP;NG2-dsRed mice.

Figure 3—figure supplement 1.

Scale bar = 10 μm.

Figure 3—figure supplement 2. Cells at NMJs express NG2 in adults but not at embryonic timepoints.

Figure 3—figure supplement 2.

(A) Immunohistochemical labeling of NG2 revealed that S100β-GFP+ cells at NMJs do not express NG2 in E18 S100β-GFP EDL muscle. (B) In the adult S100β-GFP EDL muscle, however, NG2 is detected in S100β-GFP+ cells at NMJs. Scale bar = 10 μm.
Open in a new tab

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. 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).

Figure 4.

Figure 4—figure supplement 1. The number of PSCs associated with an NMJ varies but PSC density remains constant in the EDL, soleus (SOL), and diaphragm (DIA) muscles of adult mice.

Figure 4—figure supplement 1.

(A–B) Representative images of NMJs, identified by nAChR clusters with fBTX (blue), in the (A) EDL and (B) diaphragm of adult S100β-GFP (green);NG2-dsRed (red) mice. (C) An analysis of the average number of PSCs at an NMJ shows that there are fewer PSCs at NMJs in the diaphragm as compared to the EDL and soleus. (D) An analysis of NMJs based on PSC number shows that the diaphragm has more NMJs with 1 PSC and fewer NMJs with 3 PSCs as compared to EDL and soleus. (E) When standardizing for NMJ size, the density of PSCs at NMJs is unchanged between the EDL, soleus and diaphragm. Error bar = standard error of the mean. Scale bar = 10 μm.
Open in a new tab

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).

Figure 5. Molecular analysis of S100β-GFP+;NG2-dsRed+ PSCs, S100β-GFP+ Schwan cells, and NG2-dsRed+ cells following isolation with FACS.

Figure 5.

Open in a new tab

(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.

Gene Description Proposed role Read count Log2 change vs
NG2-dsRed+		 Log2 change
vs S100β-GFP+		 Reference
Adora2a Adenosine A2a receptor Detect/modulate synaptic activity 8.1 −3.68 −2.67 (Robitaille, 1995; Rochon et al., 2001)
Adora2b Adenosine A2b receptor Detect/modulate synaptic activity 9.2 −3.16 −4.55 (Robitaille, 1995; Rochon et al., 2001)
Agrn Agrin AChR aggregation 2049.7 1.16 2.93 (Georgiou and Charlton, 1999)
Bche Butyrylcholinesterase Modulate synaptic ACh levels 7191.0 7.89 7.21 (Trachtenberg and Thompson, 1996)
Cacna1c L type Calcium channel, alpha 1 c Detect/modulate synaptic activity 14.3 −4.92 −2.10 (Morris et al., 1999)
Cacna1d L type Calcium channel, alpha 1d Detect/modulate synaptic activity 18.4 −0.42 −1.49 (Morris et al., 1999)
Cd44 CD44 antigen Mediates cell-cell interactions 1249.2 0.75 −1.22 (Woldeyesus et al., 1999)
Chrm1 Muscarinic acetylcholine receptor M1 Detect/modulate synaptic activity 14.8 n.d. 0.89 (Robitaille et al., 1997; Riethmacher et al., 1997)
Cnp 2',3'-cyclicnucleotide 3' phosphodiesterase Anchors axon terminal at NMJ 2990.2 4.23 1.66 (Personius et al., 2016)
Erbb2 Erb-b2 receptor tyrosine kinase 2 Synaptogenesis/maintenance 228.9 0.84 1.37 (Park et al., 2017; Pinard et al., 2003; Descarries et al., 1998)
Erbb3 Erb-b2 receptor tyrosine kinase 3 Synaptogenesis/maintenance 2471.3 7.05 4.46 (Park et al., 2017; Hess et al., 2007)
GAb1 GRB2-associated protein 1 Synaptic pruning 693.8 0.31 1.57 (Heredia et al., 2018)
Grm1 Glutamate receptor, metabotropic 1 Detect/modulate synaptic activity 9.2 n.d. 0.80 (Darabid et al., 2018)
Grm5 Glutamate receptor, metabotropic 5 Detect/modulate synaptic activity 38.0 n.d. 2.84 (Darabid et al., 2018)
LNX1 Ligand of numb-protein X 1 Regulate NRG1 signaling 37.5 −2.29 −0.70 (Peper et al., 1974)
MAG Myelin-associated glycoprotein Synaptogenesis/maintenance 136.0 3.12 −0.55 (Personius et al., 2016)
Mpz Myelin protein zero Synaptogenesis/maintenance 4590.7 2.54 −0.79 (Personius et al., 2016)
Nos2 Nitric oxide synthase 2, inducible Synaptogenesis/modulate synaptic activity 13.4 −2.91 −1.28 (Musarella et al., 2006)
Nos3 Nitric oxide synthase 3, endothelial cell Synaptogenesis/modulate synaptic activity 48.6 −2.69 −0.68 (Musarella et al., 2006)
P2ry1 Purinergic receptor P2Y1 Detect/modulate synaptic activity, synapse elimination 144.4 0.52 2.21 (Robitaille, 1995; De Winter et al., 2006; Feng and Ko, 2008)
P2ry2 Purinergic receptor P2Y2 Detect/modulate synaptic activity 24.0 −1.55 −1.04 (Robitaille, 1995)
P2ry10b P2Y receptor family member P2Y10b Detect/modulate synaptic activity 10.0 −1.25 −3.14 (Robitaille, 1995)
P2ry12 P2Y receptor family member P2Y12 Detect/modulate synaptic activity 273.5 n.d. 3.70 (Robitaille, 1995)
P2ry14 P2Y receptor family member P2Y14 Detect/modulate synaptic activity 13.6 −3.49 −2.06 (Robitaille, 1995)
S100b S100 protein beta Intracellular signaling 1788.3 5.34 3.12 (Reynolds and Woolf, 1992)
Sema3a Semaphorin 3a Detect/modulate synaptic activity 136.6 2.95 1.07 (Yang et al., 2001)
Tgfb1 Transforming growth factor, beta 1 AChR aggregation 173.2 −1.08 −1.90 (Petrov et al., 2014)
Open in a new tab

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		 Designation Source or
reference		 Identifiers Additional
information
Genetic reagent (M. musculus) S100b-GFP PMID:15590915 MGI:3588512 Dr. Wesley Thompson (Texas A and M)
Genetic reagent (M. musculus) NG2-dsRed PMID:18045844 MGI:3796063 Dr. Akiko Nishiyama (University of Connecticut)
Genetic reagent (M. musculus) SOD1G93A PMID:8209258 MGI:2183719 Dr. Deng (Northwestern
University)
Antibody Guinea pig polyclonal anti-NG2 PMID:19058188 Antibody Registry: AB_2572299 1:250
Antibody Alexa Fluor-488 goat polycolonal anti guinea pig Invitrogen RRID:AB_2534117 1:1000
Antibody Alexa Fluor-488 goat polyclonal anti rabbit Invitrogen Catalog# A-11008 1:1000
Software, algorithm Ingenuity Pathway Analysis Qiagen RRID:SCR_008117
Software, algorithm GraphPad Prism GraphPad RRID:SCR_002798
Software, algorithm R The R Project for Statistical Computing RRID:SCR_001905
Software, algorithm ImageJ ImageJ RRID:SCR_003070
Software, algorithm Bio-Rad CFX Manager Bio-Rad RRID:SCR_017251
Commercial assay or kit PicoPure RNA Isolation Kit ThermoFisher Catalog#KIT0204
Commercial assay or kit iScript cDNA synthesis kit Bio-Rad Catalog#1708891
Commercial assay or kit SsoAdvanced PreAmp Supermix Bio-Rad Cataolog#1725160
Commercial assay or kit iTAQ Univeral SYBR Green Supermix Bio-Rad Catalog#1725121
Chemical compound, drug Alexa Fluor-555 alpha- bungarotoxin Invitrogen Catalog#B35451
Chemical compound, drug DAPI ThermoFisher Catalog#D1306
Open in a new tab

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

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

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

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

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

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

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

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

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.

Acknowledgements

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

Funding Statement

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

Contributor Information

Gregorio Valdez, Email: gregorio_valdez@brown.edu.

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

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

Funding Information

This paper was supported by the following grants:

  • National Institutes of Health R01AG055545 to Gregorio Valdez.

  • National Institutes of Health R56AG051501 to Gregorio Valdez.

  • National Institutes of Health R21NS106313 to Gregorio Valdez.

Additional information

Competing interests

No competing interests declared.

Author contributions

Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review and editing.

Formal analysis, Investigation, Methodology, Writing - original draft, Writing - review and editing.

Data curation, Formal analysis, Writing - original draft, Writing - review and editing.

Formal analysis, Writing - review and editing.

Resources, Writing - review and editing.

Formal analysis, Methodology, Writing - original draft, Writing - review and editing.

Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing - original draft, Writing - review and editing.

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.

Additional files

Supplementary file 1. Genes with highly enriched expression in perisynaptic Schwann cells.
elife-56935-supp1.docx (101.7KB, docx)
Supplementary file 2. Primers used for cDNA preamplification and qPCR.
elife-56935-supp2.docx (16.2KB, docx)
Transparent reporting form

Data availability

RNA-Seq data has been deposited in NCBI GEO and all other data is available in the main text or the supplementary materials. The GEO accession number for this dataset is GSE152774.

The following dataset was generated:

Castro R, Taetzsch T, Vaughan SK, Godbe K, Chappell J, Settlage RE, Valdez G. 2020. Synaptic Schwann cells: specific labeling reveals unique cellular and molecular features. NCBI Gene Expression Omnibus. GSE152774

References

  1. Anders S, Pyl PT, Huber W. HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015;31:166–169. doi: 10.1093/bioinformatics/btu638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aronesty E. Comparison of sequencing utility programs. The Open Bioinformatics Journal. 2013;7:1–8. doi: 10.2174/1875036201307010001. [DOI] [Google Scholar]
  3. Astrow SH, Son YJ, Thompson WJ. Differential neural regulation of a neuromuscular junction-associated antigen in muscle fibers and schwann cells. Journal of Neurobiology. 1994;25:937–952. doi: 10.1002/neu.480250804. [DOI] [PubMed] [Google Scholar]
  4. Astrow SH, Qiang H, Ko CP. Perisynaptic schwann cells at neuromuscular junctions revealed by a novel monoclonal antibody. Journal of Neurocytology. 1998;27:667–681. doi: 10.1023/a:1006916232627. [DOI] [PubMed] [Google Scholar]
  5. Barik A, Li L, Sathyamurthy A, Xiong WC, Mei L. Schwann cells in neuromuscular junction formation and maintenance. Journal of Neuroscience. 2016;36:9770–9781. doi: 10.1523/JNEUROSCI.0174-16.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bergles DE, Jabs R, Steinhäuser C. Neuron-glia synapses in the brain. Brain Research Reviews. 2010;63:130–137. doi: 10.1016/j.brainresrev.2009.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A, Delbono O. Role of pericytes in skeletal muscle regeneration and fat accumulation. Stem Cells and Development. 2013a;22:2298–2314. doi: 10.1089/scd.2012.0647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A, Delbono O. Skeletal muscle pericyte subtypes differ in their differentiation potential. Stem Cell Research. 2013b;10:67–84. doi: 10.1016/j.scr.2012.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Boeke J. The sympathetic endformation, its synaptology, the interstitial cells, the periterminal network, and its bearing on the neurone theory discussion and critique. Cells Tissues Organs. 1949;8:18–61. doi: 10.1159/000140398. [DOI] [PubMed] [Google Scholar]
  10. Brill MS, Lichtman JW, Thompson W, Zuo Y, Misgeld T. Spatial constraints dictate glial territories at murine neuromuscular junctions. Journal of Cell Biology. 2011;195:293–305. doi: 10.1083/jcb.201108005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brockes JP, Fields KL, Raff MC. Studies on cultured rat schwann cells. I. establishment of purified populations from cultures of peripheral nerve. Brain Research. 1979;165:105–118. doi: 10.1016/0006-8993(79)90048-9. [DOI] [PubMed] [Google Scholar]
  12. Changeux JP, Kasai M, Lee CY. Use of a snake venom toxin to characterize the cholinergic receptor protein. PNAS. 1970;67:1241–1247. doi: 10.1073/pnas.67.3.1241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cook JE. Spatial properties of retinal mosaics: an empirical evaluation of some existing measures. Visual Neuroscience. 1996;13:15–30. doi: 10.1017/S0952523800007094. [DOI] [PubMed] [Google Scholar]
  14. Couteaux R. The Structure and Function of Muscle. Elsevier; 1960. [DOI] [Google Scholar]
  15. Covault J, Sanes JR. Distribution of N-CAM in synaptic and extrasynaptic portions of developing and adult skeletal muscle. The Journal of Cell Biology. 1986;102:716–730. doi: 10.1083/jcb.102.3.716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dalkin W, Taetzsch T, Valdez G. The fibular nerve injury method: a reliable assay to identify and test factors that repair neuromuscular junctions. Journal of Visualized Experiments. 2016;11:54186. doi: 10.3791/54186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Darabid H, Arbour D, Robitaille R. Glial cells decipher synaptic competition at the mammalian neuromuscular junction. Journal of Neuroscience. 2013;33:1297–1313. doi: 10.1523/JNEUROSCI.2935-12.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Darabid H, Perez-Gonzalez AP, Robitaille R. Neuromuscular synaptogenesis: coordinating partners with multiple functions. Nature Reviews Neuroscience. 2014;15:703–718. doi: 10.1038/nrn3821. [DOI] [PubMed] [Google Scholar]
  19. Darabid H, St-Pierre-See A, Robitaille R. Purinergic-Dependent glial regulation of synaptic plasticity of competing terminals and synapse elimination at the neuromuscular junction. Cell Reports. 2018;25:2070–2082. doi: 10.1016/j.celrep.2018.10.075. [DOI] [PubMed] [Google Scholar]
  20. De Winter F, Vo T, Stam FJ, Wisman LA, Bär PR, Niclou SP, van Muiswinkel FL, Verhaagen J. The expression of the chemorepellent semaphorin 3A is selectively induced in terminal schwann cells of a subset of neuromuscular synapses that display limited anatomical plasticity and enhanced vulnerability in motor neuron disease. Molecular and Cellular Neuroscience. 2006;32:102–117. doi: 10.1016/j.mcn.2006.03.002. [DOI] [PubMed] [Google Scholar]
  21. Deloulme JC, Raponi E, Gentil BJ, Bertacchi N, Marks A, Labourdette G, Baudier J. Nuclear expression of S100B in oligodendrocyte progenitor cells correlates with differentiation toward the oligodendroglial lineage and modulates oligodendrocytes maturation. Molecular and Cellular Neuroscience. 2004;27:453–465. doi: 10.1016/j.mcn.2004.07.008. [DOI] [PubMed] [Google Scholar]
  22. Descarries LM, Cai S, Robitaille R, Josephson EM, Morest DK. Localization and characterization of nitric oxide synthase at the frog neuromuscular junction. Journal of Neurocytology. 1998;27:829–840. doi: 10.1023/a:1006907531778. [DOI] [PubMed] [Google Scholar]
  23. Fatt P, Katz B. The electric activity of the motor end-plate. Proc. R. Soc. London. Ser. B, Biol. Sci. 1952;140:183–186. doi: 10.1098/rspb.1952.0055. [DOI] [PubMed] [Google Scholar]
  24. Feng Z, Ko CP. Schwann cells promote synaptogenesis at the neuromuscular junction via transforming growth factor-beta1. Journal of Neuroscience. 2008;28:9599–9609. doi: 10.1523/JNEUROSCI.2589-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Filous AR, Tran A, Howell CJ, Busch SA, Evans TA, Stallcup WB, Kang SH, Bergles DE, Lee SI, Levine JM, Silver J. Entrapment via synaptic-like connections between NG2 proteoglycan+ cells and dystrophic axons in the lesion plays a role in regeneration failure after spinal cord injury. The Journal of Neuroscience. 2014;34:16369–16384. doi: 10.1523/JNEUROSCI.1309-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Fox MA, Umemori H. Seeking long-term relationship: axon and target communicate to organize synaptic differentiation. Journal of Neurochemistry. 2006;97:1215–1231. doi: 10.1111/j.1471-4159.2006.03834.x. [DOI] [PubMed] [Google Scholar]
  27. Georgiou J, Charlton MP. Non-myelin-forming perisynaptic schwann cells express protein zero and myelin-associated glycoprotein. Glia. 1999;27:101–109. doi: 10.1002/(SICI)1098-1136(199908)27:2&#x0003c;101::AID-GLIA1&#x0003e;3.0.CO;2-H. [DOI] [PubMed] [Google Scholar]
  28. Godfrey EW, Nitkin RM, Wallace BG, Rubin LL, McMahan UJ. Components of Torpedo electric organ and muscle that cause aggregation of acetylcholine receptors on cultured muscle cells. The Journal of Cell Biology. 1984;99:615–627. doi: 10.1083/jcb.99.2.615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Gorlewicz A, Wlodarczyk J, Wilczek E, Gawlak M, Cabaj A, Majczynski H, Nestorowicz K, Herbik MA, Grieb P, Slawinska U, Kaczmarek L, Wilczynski GM. CD44 is expressed in non-myelinating schwann cells of the adult rat, and may play a role in neurodegeneration-induced glial plasticity at the neuromuscular junction. Neurobiology of Disease. 2009;34:245–258. doi: 10.1016/j.nbd.2009.01.011. [DOI] [PubMed] [Google Scholar]
  30. Griffin JW, Thompson WJ. Biology and pathology of nonmyelinating schwann cells. Glia. 2008;56:1518–1531. doi: 10.1002/glia.20778. [DOI] [PubMed] [Google Scholar]
  31. Gurney ME, Pu H, Chiu AY, Dal Canto MC, Polchow CY, Alexander DD, Caliendo J, Hentati A, Kwon YW, Deng HX. Motor neuron degeneration in mice that express a human cu,zn superoxide dismutase mutation. Science. 1994;264:1772–1775. doi: 10.1126/science.8209258. [DOI] [PubMed] [Google Scholar]
  32. Hachem S, Aguirre A, Vives V, Marks A, Gallo V, Legraverend C. Spatial and temporal expression of S100B in cells of oligodendrocyte lineage. Glia. 2005;51:81–97. doi: 10.1002/glia.20184. [DOI] [PubMed] [Google Scholar]
  33. Heredia DJ, Feng CY, Agarwal A, Nennecker K, Hennig GW, Gould TW. Postnatal restriction of Activity-Induced Ca2+ Responses to Schwann Cells at the Neuromuscular Junction Are Caused by the Proximo-Distal Loss of Axonal Synaptic Vesicles during Development. The Journal of Neuroscience. 2018;38:8650–8665. doi: 10.1523/JNEUROSCI.0956-18.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hess DM, Scott MO, Potluri S, Pitts EV, Cisterni C, Balice-Gordon RJ. Localization of TrkC to schwann cells and effects of neurotrophin-3 signaling at neuromuscular synapses. The Journal of Comparative Neurology. 2007;501:465–482. doi: 10.1002/cne.21163. [DOI] [PubMed] [Google Scholar]
  35. Heuser JE, Reese TS. Evidence for recycling of synaptic vesicle membrane during transmitter release at the frog neuromuscular junction. The Journal of Cell Biology. 1973;57:315–344. doi: 10.1083/jcb.57.2.315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Jahromi BS, Robitaille R, Charlton MP. Transmitter release increases intracellular calcium in Perisynaptic schwann cells in situ. Neuron. 1992;8:1069–1077. doi: 10.1016/0896-6273(92)90128-Z. [DOI] [PubMed] [Google Scholar]
  37. Jäkel S, Dimou L. Glial cells and their function in the adult brain: a journey through the history of their ablation. Frontiers in Cellular Neuroscience. 2017;11:24. doi: 10.3389/fncel.2017.00024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Jennings CG, Dyer SM, Burden SJ. Muscle-specific trk-related receptor with a kringle domain defines a distinct class of receptor tyrosine kinases. PNAS. 1993;90:2895–2899. doi: 10.1073/pnas.90.7.2895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Kang H, Tian L, Son YJ, Zuo Y, Procaccino D, Love F, Hayworth C, Trachtenberg J, Mikesh M, Sutton L, Ponomareva O, Mignone J, Enikolopov G, Rimer M, Thompson W. Regulation of the intermediate filament protein nestin at rodent neuromuscular junctions by innervation and activity. Journal of Neuroscience. 2007;27:5948–5957. doi: 10.1523/JNEUROSCI.0621-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Katz B, Miledi R. Input-output relation of a single synapse. Nature. 1966;212:1242–1245. doi: 10.1038/2121242a0. [DOI] [PubMed] [Google Scholar]
  41. Katz B, Miledi R. A study of synaptic transmission in the absence of nerve impulses. The Journal of Physiology. 1967;192:407–436. doi: 10.1113/jphysiol.1967.sp008307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Kim D, Paggi JM, Park C, Bennett C, Salzberg SL. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nature Biotechnology. 2019;37:907–915. doi: 10.1038/s41587-019-0201-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ko CP, Robitaille R. Perisynaptic schwann cells at the neuromuscular synapse: adaptable, multitasking glial cells. Cold Spring Harbor Perspectives in Biology. 2015;7:a020503. doi: 10.1101/cshperspect.a020503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Koirala S, Qiang H, Ko CP. Reciprocal interactions between perisynaptic schwann cells and regenerating nerve terminals at the frog neuromuscular junction. Journal of Neurobiology. 2000;44:343–360. doi: 10.1002/1097-4695(20000905)44:3&#x0003c;343::AID-NEU5&#x0003e;3.0.CO;2-O. [DOI] [PubMed] [Google Scholar]
  45. Kong Y. Btrim: a fast, lightweight adapter and quality trimming program for next-generation sequencing technologies. Genomics. 2011;98:152–153. doi: 10.1016/j.ygeno.2011.05.009. [DOI] [PubMed] [Google Scholar]
  46. Lee YI, Thompson WJ, Harlow ML. Schwann cells participate in synapse elimination at the developing neuromuscular junction. Current Opinion in Neurobiology. 2017;47:176–181. doi: 10.1016/j.conb.2017.10.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Love FM, Thompson WJ. Schwann cells proliferate at rat neuromuscular junctions during development and regeneration. The Journal of Neuroscience. 1998;18:9376–9385. doi: 10.1523/JNEUROSCI.18-22-09376.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Lwebuga-Mukasa JS, Lappi S, Taylor P. Molecular forms of acetylcholinesterase from Torpedo californica: their relationship to synaptic membranes. Biochemistry. 1976;15:1425–1434. doi: 10.1021/bi00652a012. [DOI] [PubMed] [Google Scholar]
  50. Matthias K, Kirchhoff F, Seifert G, Hüttmann K, Matyash M, Kettenmann H, Steinhäuser C. Segregated expression of AMPA-type glutamate receptors and glutamate transporters defines distinct astrocyte populations in the mouse Hippocampus. The Journal of Neuroscience. 2003;23:1750–1758. doi: 10.1523/JNEUROSCI.23-05-01750.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Miledi R, Slater CR. Electrophysiology and electron-microscopy of rat neuromuscular junctions after nerve degeneration. Proc. R. Soc. London. Ser. B. Biol. Sci. 1968;169:289–306. doi: 10.1098/rspb.1968.0012. [DOI] [PubMed] [Google Scholar]
  52. Miledi R, Slater CR. On the degeneration of rat neuromuscular junctions after nerve section. The Journal of Physiology. 1970;207:507–528. doi: 10.1113/jphysiol.1970.sp009076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Moloney EB, de Winter F, Verhaagen J. ALS as a distal axonopathy: molecular mechanisms affecting neuromuscular junction stability in the presymptomatic stages of the disease. Frontiers in Neuroscience. 2014;8:252. doi: 10.3389/fnins.2014.00252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Morris JK, Lin W, Hauser C, Marchuk Y, Getman D, Lee KF. Rescue of the cardiac defect in ErbB2 mutant mice reveals essential roles of ErbB2 in peripheral nervous system development. Neuron. 1999;23:273–283. doi: 10.1016/S0896-6273(00)80779-5. [DOI] [PubMed] [Google Scholar]
  55. Moshrefi-Ravasdjani B, Dublin P, Seifert G, Jennissen K, Steinhäuser C, Kafitz KW, Rose CR. Changes in the proliferative capacity of NG2 cell subpopulations during postnatal development of the mouse Hippocampus. Brain Structure and Function. 2017;222:831–847. doi: 10.1007/s00429-016-1249-2. [DOI] [PubMed] [Google Scholar]
  56. Mozer BA, Sandstrom DJ. Drosophila neuroligin 1 regulates synaptic growth and function in response to activity and phosphoinositide-3-kinase. Molecular and Cellular Neuroscience. 2012;51:89–100. doi: 10.1016/j.mcn.2012.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Musarella M, Alcaraz G, Caillol G, Boudier JL, Couraud F, Autillo-Touati A. Expression of Nav1.6 sodium channels by Schwann cells at neuromuscular junctions: role in the motor endplate disease phenotype. Glia. 2006;53:13–23. doi: 10.1002/glia.20252. [DOI] [PubMed] [Google Scholar]
  58. Nitkin RM, Smith MA, Magill C, Fallon JR, Yao YM, Wallace BG, McMahan UJ. Identification of Agrin, a synaptic organizing protein from Torpedo electric organ. The Journal of Cell Biology. 1987;105:2471–2478. doi: 10.1083/jcb.105.6.2471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. O'Malley JP, Waran MT, Balice-Gordon RJ. In vivo observations of terminal schwann cells at normal, denervated, and reinnervated mouse neuromuscular junctions. Journal of Neurobiology. 1999;38:270–286. doi: 10.1002/(SICI)1097-4695(19990205)38:2&#x0003c;270::AID-NEU9&#x0003e;3.0.CO;2-F. [DOI] [PubMed] [Google Scholar]
  60. Park SY, Jang SY, Shin YK, Jung DK, Yoon BA, Kim JK, Jo YR, Lee HJ, Park HT. The scaffolding protein, Grb2-associated Binder-1, in skeletal muscles and terminal schwann cells regulates postnatal neuromuscular synapse maturation. Experimental Neurobiology. 2017;26:141–150. doi: 10.5607/en.2017.26.3.141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Peng F, Yao H, Bai X, Zhu X, Reiner BC, Beazely M, Funa K, Xiong H, Buch S. Platelet-derived growth factor-mediated induction of the synaptic plasticity gene arc/Arg3.1. Journal of Biological Chemistry. 2010;285:21615–21624. doi: 10.1074/jbc.M110.107003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Peper K, Dreyer F, Sandri C, Akert K, Moor H. Structure and ultrastructure of the frog motor endplate a freeze-etching study. Cell and Tissue Research. 1974;149:437–455. doi: 10.1007/BF00223024. [DOI] [PubMed] [Google Scholar]
  63. Perez VJ, Moore BW. Wallerian degeneration in rabbit tibial nerve: changes in amounts of the S-100 protein. Journal of Neurochemistry. 1968;15:971–977. doi: 10.1111/j.1471-4159.1968.tb11640.x. [DOI] [PubMed] [Google Scholar]
  64. Personius KE, Slusher BS, Udin SB. Neuromuscular NMDA receptors modulate developmental synapse elimination. The Journal of Neuroscience. 2016;36:8783–8789. doi: 10.1523/JNEUROSCI.1181-16.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Petrov KA, Girard E, Nikitashina AD, Colasante C, Bernard V, Nurullin L, Leroy J, Samigullin D, Colak O, Nikolsky E, Plaud B, Krejci E. Schwann cells sense and control acetylcholine spillover at the neuromuscular junction by α7 nicotinic receptors and butyrylcholinesterase. Journal of Neuroscience. 2014;34:11870–11883. doi: 10.1523/JNEUROSCI.0329-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Pinard A, Lévesque S, Vallée J, Robitaille R. Glutamatergic modulation of synaptic plasticity at a PNS vertebrate cholinergic synapse. European Journal of Neuroscience. 2003;18:3241–3250. doi: 10.1111/j.1460-9568.2003.03028.x. [DOI] [PubMed] [Google Scholar]
  67. Platel JC, Gordon V, Heintz T, Bordey A. GFAP-GFP neural progenitors are antigenically homogeneous and anchored in their enclosed mosaic niche. Glia. 2009;57:66–78. doi: 10.1002/glia.20735. [DOI] [PubMed] [Google Scholar]
  68. Porter S, Froehner SC. Characterization and localization of the mr = 43,000 proteins associated with acetylcholine receptor-rich membranes. The Journal of Biological Chemistry. 1983;258:10034–10040. [PubMed] [Google Scholar]
  69. Rafuse VF, Polo-Parada L, Landmesser LT. Structural and functional alterations of neuromuscular junctions in NCAM-deficient mice. The Journal of Neuroscience. 2000;20:6529–6539. doi: 10.1523/JNEUROSCI.20-17-06529.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Ranaivoson FM, Turk LS, Ozgul S, Kakehi S, von Daake S, Lopez N, Trobiani L, De Jaco A, Denissova N, Demeler B, Özkan E, Montelione GT, Comoletti D. A proteomic screen of neuronal Cell-Surface molecules reveals IgLONs as structurally conserved interaction modules at the synapse. Structure. 2019;27:893–906. doi: 10.1016/j.str.2019.03.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Reddy LV, Koirala S, Sugiura Y, Herrera AA, Ko CP. Glial cells maintain synaptic structure and function and promote development of the neuromuscular junction in vivo. Neuron. 2003;40:563–580. doi: 10.1016/S0896-6273(03)00682-2. [DOI] [PubMed] [Google Scholar]
  72. Reese BE, Keeley PW. Design principles and developmental mechanisms underlying retinal mosaics. Biological Reviews. 2015;90:854–876. doi: 10.1111/brv.12139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Reist NE, Smith SJ. Neurally evoked calcium transients in terminal schwann cells at the neuromuscular junction. PNAS. 1992;89:7625–7629. doi: 10.1073/pnas.89.16.7625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Reynolds ML, Woolf CJ. Terminal schwann cells elaborate extensive processes following denervation of the motor endplate. Journal of Neurocytology. 1992;21:50–66. doi: 10.1007/BF01206897. [DOI] [PubMed] [Google Scholar]
  75. Riethmacher D, Sonnenberg-Riethmacher E, Brinkmann V, Yamaai T, Lewin GR, Birchmeier C. Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature. 1997;389:725–730. doi: 10.1038/39593. [DOI] [PubMed] [Google Scholar]
  76. Robertson JD. The ultrastructure of a reptilian myoneural junction. The Journal of Biophysical and Biochemical Cytology. 1956;2:381–394. doi: 10.1083/jcb.2.4.381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Robitaille R. Purinergic receptors and their activation by endogenous purines at Perisynaptic glial cells of the frog neuromuscular junction. The Journal of Neuroscience. 1995;15:7121–7131. doi: 10.1523/JNEUROSCI.15-11-07121.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Robitaille R, Bourque MJ, Vandaele S. Localization of L-type Ca2+ channels at Perisynaptic glial cells of the frog neuromuscular junction. The Journal of Neuroscience. 1996;16:148–158. doi: 10.1523/JNEUROSCI.16-01-00148.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Robitaille R, Jahromi BS, Charlton MP. Muscarinic Ca2+ responses resistant to muscarinic antagonists at Perisynaptic schwann cells of the frog neuromuscular junction. The Journal of Physiology. 1997;504:337–347. doi: 10.1111/j.1469-7793.1997.337be.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Rochon D, Rousse I, Robitaille R. Synapse-glia interactions at the mammalian neuromuscular junction. The Journal of Neuroscience. 2001;21:3819–3829. doi: 10.1523/JNEUROSCI.21-11-03819.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Sanes JR, Lichtman JW. Development of the vertebrate neuromuscular junction. Annual Review of Neuroscience. 1999;22:389–442. doi: 10.1146/annurev.neuro.22.1.389. [DOI] [PubMed] [Google Scholar]
  82. Sanes JR, Lichtman JW. Induction, assembly, maturation and maintenance of a postsynaptic apparatus. Nature Reviews Neuroscience. 2001;2:791–805. doi: 10.1038/35097557. [DOI] [PubMed] [Google Scholar]
  83. Sealock R, Murnane AA, Paulin D, Froehner SC. Immunochemical identification of desmin in Torpedo postsynaptic membranes and at the rat neuromuscular junction. Synapse. 1989;3:315–324. doi: 10.1002/syn.890030404. [DOI] [PubMed] [Google Scholar]
  84. Shapiro L, Love J, Colman DR. Adhesion molecules in the nervous system: structural insights into function and diversity. Annual Review of Neuroscience. 2007;30:451–474. doi: 10.1146/annurev.neuro.29.051605.113034. [DOI] [PubMed] [Google Scholar]
  85. Smith IW, Mikesh M, Lee Y, Thompson WJ. Terminal schwann cells participate in the competition underlying neuromuscular synapse elimination. Journal of Neuroscience. 2013;33:17724–17736. doi: 10.1523/JNEUROSCI.3339-13.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Sobel A, Weber M, Changeux JP. Large-scale purification of the acetylcholine-receptor protein in its membrane-bound and detergent-extracted forms from Torpedo marmorata electric organ. European Journal of Biochemistry. 1977;80:215–224. doi: 10.1111/j.1432-1033.1977.tb11874.x. [DOI] [PubMed] [Google Scholar]
  87. Sobel A, Hofler J, Heidmann T, Changeux JP. Structural and functional properties of the acetylcholine regulator. Advances in Cytopharmacology. 1979;3:191–196. [PubMed] [Google Scholar]
  88. Son YJ, Thompson WJ. Schwann cell processes guide regeneration of peripheral axons. Neuron. 1995;14:125–132. doi: 10.1016/0896-6273(95)90246-5. [DOI] [PubMed] [Google Scholar]
  89. Stallcup WB. The NG2 antigen, a putative lineage marker: immunofluorescent localization in primary cultures of rat brain. Developmental Biology. 1981;83:154–165. doi: 10.1016/S0012-1606(81)80018-8. [DOI] [PubMed] [Google Scholar]
  90. Trachtenberg JT, Thompson WJ. Schwann cell apoptosis at developing neuromuscular junctions is regulated by glial growth factor. Nature. 1996;379:174–177. doi: 10.1038/379174a0. [DOI] [PubMed] [Google Scholar]
  91. Trachtenberg JT, Thompson WJ. Nerve terminal withdrawal from rat neuromuscular junctions induced by neuregulin and schwann cells. The Journal of Neuroscience. 1997;17:6243–6255. doi: 10.1523/JNEUROSCI.17-16-06243.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  92. Vives V, Alonso G, Solal AC, Joubert D, Legraverend C. Visualization of S100B-positive neurons and Glia in the central nervous system of EGFP transgenic mice. The Journal of Comparative Neurology. 2003;457:404–419. doi: 10.1002/cne.10552. [DOI] [PubMed] [Google Scholar]
  93. Wassle H, Riemann HJ. The mosaic of Nhtman,erve cells in the mammalian retina. Proc. R. Soc. London - Biol. Sci. 1978;200:441–461. doi: 10.1098/rspb.1978.0026. [DOI] [PubMed] [Google Scholar]
  94. Woldeyesus MT, Britsch S, Riethmacher D, Xu L, Sonnenberg-Riethmacher E, Abou-Rebyeh F, Harvey R, Caroni P, Birchmeier C. Peripheral nervous system defects in erbB2 mutants following genetic rescue of heart development. Genes & Development. 1999;13:2538–2548. doi: 10.1101/gad.13.19.2538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  95. Wright MC, Potluri S, Wang X, Dentcheva E, Gautam D, Tessler A, Wess J, Rich MM, Son YJ. Distinct muscarinic acetylcholine receptor subtypes contribute to stability and growth, but not compensatory plasticity, of neuromuscular synapses. Journal of Neuroscience. 2009;29:14942–14955. doi: 10.1523/JNEUROSCI.2276-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  96. Yang JF, Cao G, Koirala S, Reddy LV, Ko CP. Schwann cells express active agrin and enhance aggregation of acetylcholine receptors on muscle fibers. The Journal of Neuroscience. 2001;21:9572–9584. doi: 10.1523/JNEUROSCI.21-24-09572.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  97. Young P, Nie J, Wang X, McGlade CJ, Rich MM, Feng G. LNX1 is a perisynaptic schwann cell specific E3 ubiquitin ligase that interacts with ErbB2. Molecular and Cellular Neuroscience. 2005;30:238–248. doi: 10.1016/j.mcn.2005.07.015. [DOI] [PubMed] [Google Scholar]
  98. Zhu X, Bergles DE, Nishiyama A. NG2 cells generate both oligodendrocytes and gray matter astrocytes. Development. 2008;135:145–157. doi: 10.1242/dev.004895. [DOI] [PubMed] [Google Scholar]
  99. Zuo Y, Lubischer JL, Kang H, Tian L, Mikesh M, Marks A, Scofield VL, Maika S, Newman C, Krieg P, Thompson WJ. Fluorescent proteins expressed in mouse transgenic lines mark subsets of Glia, neurons, macrophages, and dendritic cells for vital examination. Journal of Neuroscience. 2004;24:10999–11009. doi: 10.1523/JNEUROSCI.3934-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision letter

Editor: Beth Stevens1
Reviewed by: Chien-Ping Ko2

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.

eLife. 2020 Jun 25;9:e56935. doi: 10.7554/eLife.56935.sa2

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.

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Data Citations

    1. Castro R, Taetzsch T, Vaughan SK, Godbe K, Chappell J, Settlage RE, Valdez G. 2020. Synaptic Schwann cells: specific labeling reveals unique cellular and molecular features. NCBI Gene Expression Omnibus. GSE152774 [DOI] [PMC free article] [PubMed]

    Supplementary Materials

    Supplementary file 1. Genes with highly enriched expression in perisynaptic Schwann cells.
    elife-56935-supp1.docx (101.7KB, docx)
    Supplementary file 2. Primers used for cDNA preamplification and qPCR.
    elife-56935-supp2.docx (16.2KB, docx)
    Transparent reporting form
    elife-56935-transrepform.docx (246.1KB, docx)

    Data Availability Statement

    RNA-Seq data has been deposited in NCBI GEO and all other data is available in the main text or the supplementary materials. The GEO accession number for this dataset is GSE152774.

    The following dataset was generated:

    Castro R, Taetzsch T, Vaughan SK, Godbe K, Chappell J, Settlage RE, Valdez G. 2020. Synaptic Schwann cells: specific labeling reveals unique cellular and molecular features. NCBI Gene Expression Omnibus. GSE152774

    Articles from eLife are provided here courtesy of eLife Sciences Publications, Ltd

    RESOURCES