Axo‐glial interaction in the injured PNS

1 INTRODUCTION

The peripheral nervous system (PNS) is characterized by an intimate relationship between axons and the resident peripheral glial cells, the Schwann cells. Schwann cells engulf their associated axons and form an isolating myelin sheath around nerve fibers larger than 1µm in size. Both cell types depend on each other for long-term survival and integrity and a disturbance of this fine-tuned interplay causes peripheral nerve diseases, which comprise a heterogeneous group of acute and chronic disorders with different acquired and hereditary origins. Research on acute nerve injury and hereditary neuropathies, collectively termed Charcot-Marie-Tooth diseases, has greatly informed our understanding of Schwann cell biology and plasticity. Acute peripheral nerve injury is induced by neurotmesis (nerve cut) or axonotmesis (nerve crush) and results in a sequence of events referred to as Wallerian degeneration, which comprises the degeneration of the axon distal to the injury site and the subsequent clearance of axonal debris (Waller, 1850). Importantly, in contrast to axons, Schwann cells do not degenerate after axonal injury but transdifferentiate into specific repair cells, which essentially facilitate axonal regrowth and remyelination of the injured nerve (Jessen & Mirsky, 2016). The repair of a nerve cut, however, first of all, requires the bridging of the gap between proximal and distal nerve stumps, a concerted process that is realized by the shared labor of macrophages, endothelial cells, fibroblasts and Schwann cells (Cattin & Lloyd, 2016).

The generation of the repair Schwann cell encompasses the transdifferentiation into an immature like state and the concomitant upregulation of specific transcription factors and genes that support repair (for reviews on this topic see Jessen & Arthur-Farraj, 2019; Jessen & Mirsky, 2016). This switch in the phenotypic state of Schwann cells is referred to as Schwann cell adaptive cellular reprogramming (Arthur-Farraj et al., 2012; Jessen & Arthur-Farraj, 2019; Jessen, Mirsky, & Arthur-Farraj, 2015; Masaki et al., 2013).

Schwann cells also activate the innate immune system and contribute to the removal of myelin debris via a form of selective autophagy (Brosius Lutz et al., 2017; Gomez-Sanchez et al., 2015; Jang et al., 2016). After injury-induced reprogramming, repair Schwann cells undergo proliferation and subsequently align in so-called Bands of Büngner, which provide the guidance structures for regrowing axons (Jessen & Mirsky, 2008, 2016). Notably, Schwann cells substantially elongate and branch to form repair Schwann cells, but shorten radically on remyelination (Gomez-Sanchez et al., 2017). Indeed, remyelinated axons suffer from a strongly reduced internodal length and thinner myelin sheaths, reflecting the disparity between peripheral nerve development and regeneration. Although the individual parameters that determine the difference between these two processes remain only partially understood, the lack of concomitant body growth, as well as an altered axon-glia signaling, may likely constitute contributing factors.

While Wallerian degeneration and the appearance of a Schwann cell repair phenotype are classically induced by acute nerve injury, Wallerian-like mechanisms are also a feature of “dying back” axonal degeneration in peripheral neuropathies (Coleman, 2005; Conforti, Gilley, & Coleman, 2014). Moreover, a dysdifferentiation of Schwann cells with a molecular phenotype similar to repair Schwann cells has also been shown to be a feature of demyelinating peripheral neuropathies such as CMT1A and CMT1B disease (D’Antonio et al., 2013; Fledrich et al., 2014; Martini, Klein, & Groh, 2013; Patzkó et al., 2012). Here, Schwann cells within a morphologically intact myelinating Schwann cell–axon unit acquire a repair-like phenotype (Fledrich et al., 2014; Hutton et al., 2011; Klein, Groh, Wettmarshausen, & Martini, 2014). Hence, not only Wallerian degeneration and a de facto loss of axonal contact induce an injury response in Schwann cells. In contrast, functional denervation due to an endogenous Schwann cell defect may elicit the activation of a repair phenotype in Schwann cells of peripheral neuropathies such as CMT1A disease. Taken together, the Schwann cell repair response represents a promising candidate for a common glial pathomechanism in different peripheral nerve diseases and future research may provide more information on whether this phenotype also constitutes a potential therapeutic target for these diseases.

The present review focusses on bidirectional axon-glia interactions in acute peripheral nerve injury as well as in peripheral neuropathies. For the important function of intracellular signaling pathways, transcription factors as well as the role of the immune system we refer to seminal recent reviews on this topic (Boerboom, Dion, Chariot, & Franzen, 2017; Castelnovo et al., 2017; Jessen & Arthur-Farraj, 2019; Jessen & Mirsky, 2019; Klein & Martini, 2016).

2 ACUTE NERVE INJURY 2.1 Axonal breakdown and Schwann cell transdifferentiation

Acute axonal injury to peripheral nerves, either after nerve crush or cut, results in the activation of a series of molecular events that lead to the degradation of the distal nerve stump during the first few days after injury. These include the fragmentation of axons and dissolution of their neurofilament scaffolds, the conversion of the myelinating and non-myelinating Schwann cells to repair Schwann cells, and the recruitment of inflammatory cells including macrophages to clear myelin sheaths and axon debris (Figure 1). In addition, in injuries such as nerve transection, where the basal lamina and connective tissue are completely disrupted, a nerve bridge of new tissue is formed to re-join the two nerve stumps. This profound remodeling of the distal injury stump is necessary to create an environment that is conducive to axonal regeneration.

image

Reciprocal interactions between Schwann cells and axons after physical lesion. Physical trauma to nerves results in radical changes in the nerve environment in preparation for the re-innervation of target organs. In the early Wallerian degeneration stage, there is a profound remodeling of the distal injury stumps, whereby the axons fragment and disintegrate, and the myelinating and non-myelinating Schwann cells convert to repair Schwann cells. The axon-glia interactions also undergo extensive changes, with Schwann cells driving axonal fragmentation and clearance of myelin debris initially, and then providing structural and metabolic support to the regrowing axons. After target re-innervation, axonal signals, including NRG1, activate ErbB-receptors in Schwann cells to drive axonal remyelination

Injury to the nerve subdivides the fiber tracts into two segments: a distal segment, in which the axons are separated from the neuron soma, and a proximal segment, in which the axons still remain attached to the cell body. The cell membrane and cytoskeleton of the injured axons are disrupted, which induce a stress response that leads to a resealing of the membrane and cytoskeleton remodeling. Initially, there is a similar acute phase of axonal degeneration followed by slow axonal retraction at the cut sites of both proximal and distal segments (Girouard et al., 2018; Kerschensteiner, Schwab, Lichtman, & Misgeld, 2005; Knoferle et al., 2010). However, the subsequent response of both segments then differs remarkably. In the distal segments, there is a rapid disintegration of the injured axons, with most of the axons broken down into small fragments (Beirowski et al., 2005). Within the last decade, substantial progress has been made in the elucidation of the molecular control of axonal breakdown (Coleman & Höke, 2020). For instance, SARM1, a Toll-like receptor adaptor protein, has been identified as a major effector of this response in the injured axons and is required for the normal, rapid rate of Wallerian degeneration(Coleman & Höke, 2020; Gerdts, Brace, Sasaki, DiAntonio, & Milbrandt, 2015; Gerdts, Summers, Milbrandt, & DiAntonio, 2016).

Whilst the initiation of axon degeneration appears to be mainly axon intrinsic, Schwann cells actively participate in the subsequent steps of axonal fragmentation and the clearance of axonal debris (Catenaccio et al., 2017; Raff, Whitmore, & Finn, 2002; Rosenberg, Isaacman-Beck, Franzini-Armstrong, & Granato, 2014; Rosenberg, Wolman, Franzini-Armstrong, & Granato, 2012; Vaquié et al., 2019; Villegas et al., 2012). Indeed, studies of acute axonal degeneration in mice revealed that axonal injury initiates a response in Schwann cells which results in the fragmentation of their associated axons by the actin-rich cytoplasmic spirals of the Schmidt-Lanterman incisures (Catenaccio et al., 2017). At 2 and 3 days post-injury in mice, Schmidt-Lanterman incisures were found to increase in size and to protrude into the axon creating an axonal segment encapsulated within the inter incisure myelin fragment (Catenaccio et al., 2017). Molecularly, the authors found Schwann cell-mediated early axon degeneration to overlap with cellular cytokinesis, and termed the respective process “axokinesis” (Catenaccio et al., 2017). In line with this, Vaquié et al. (2019) could demonstrate that Schwann cells could promote the formation of strings of actin spheres along unfragmented axons, which lead to a constriction/expansion of axons until their complete disintegration. Notably, the authors found that this process was initiated by a local translation of Plgf in the injured axons that activated the VEGFR1 receptor in Schwann cells and the subsequent formation of the actin spheres. These studies represent important examples of the bi-directional interaction of axons and Schwann cells in the early remodeling of the nerve after injury. Similarly, recent evidence suggests that in cases of mild injury such as crush injury, not all axons are axotomized in the distal part, and that the surviving damaged axons are removed by natural killer cells that invade the nerves shortly after injury (Davies et al., 2019). In the proximal segments, following the initial acute phase of axonal degeneration, there is a restructuring of the cytoskeletal components, such as actin and microtubules that support the formation of a new growth cone and drive axon extension (Bradke, Fawcett, & Spira, 2012; Girouard et al., 2018), (see the section below on axonal regeneration and remyelination).

Next to distal axonal breakdown and the formation of a growth cone proximally, nerve injury results in a radical change in the signaling environment of the Schwann cells. They lose contact with key signals from axons, as these degenerate in the injured stump, and become subsequently exposed to bioactive molecules from invading immune cells. This leads to the conversion of the myelinating and non-myelinating Schwann cells to a distinct alternative cell type, the repair Schwann cell. These specialized cells have multiple cell-autonomous and non-cell-autonomous roles, which are essential for orchestrating the regenerative response (Jessen & Mirsky, 2019; Stierli et al., 2018).

While the morphological fragmentation of axons only takes places after a first latent phase (~1,5–2 days in rodents), the first molecular changes in Schwann cells can be already detected within a few hours after injury (Bosse, Hasenpusch-Theil, Küry, & Müller, 2006; Gerdts et al., 2016; Jessen & Mirsky, 2008, 2019; Stassart et al., 2013). Hence, activation of Schwann cells by injured axons is highly likely, though the mechanisms still remain obscure. A potential candidate that may contribute to the activation of the repair program in Schwann cells is the Notch signaling pathway with its ligands Jagged 1 and 2, which are induced within 1 to 2 days after nerve cut in injured Schwann cells (Woodhoo et al., 2009). The activation of the Notch pathway induces demyelination in vitro and the absence of Notch signaling in vivo reduces the rate of demyelination and delays Schwann cell dedifferentiation (Woodhoo et al., 2009). However, the cellular source of the Notch ligands Jagged and Delta remain to be determined and might include Schwann cells, axons, and macrophages (Woodhoo et al., 2009). Similar to Notch signaling, ErbB receptor activation has also been shown to occur rapidly after nerve injury (Carroll, Miller, Frohnert, Kim, & Corbett, 1997; Cohen, Yachnis, Arai, Davis, & Scherer, 1992; Guertin, Zhang, Mak, Alberta, & Kim, 2005; Kwon et al., 1997), although its ligand, axonal NRG1, does not seem to play a major role in Schwann cell dedifferentiation, at least as shown by the lack of effect upon its ablation. Instead, axonal NRG1 is important later during nerve regeneration (see below; Fricker et al., 2011, 2013; Stassart et al., 2013). Notably, the induced activation of Raf/ERK, one of the major pathways downstream of ErbB receptor activation, in Schwann cells of non-injured mice has been shown to activate a Schwann cell phenotype in the adult nerve, which shares important features with the denervated Schwann cell response after acute nerve injury (Harrisingh et al., 2004; Napoli et al., 2012). Interestingly, this phenotype proved reversible upon the switch-off of the ERK signal, demonstrating that Schwann cells are very sensitive to altered activation of MEK/ERK signaling (Cervellini et al., 2018; Napoli et al., 2012). However, which signals are upstream of the Raf/ERK pathway upon acute nerve injury remain to be determined.

The conversion of myelinating and non-myelinating Schwann cells to the repair Schwann cell is associated with the reversal of their differentiation program. For example, an immediate response of myelin Schwann cells is the downregulation of genes encoding key myelin transcription factors, including Egr2, and structural proteins, including MPZ, MBP and Periaxin, and conversely upregulation of other genes that are normally expressed in non-myelinating or developing Schwann cells, including L1, NCAM, GFAP and P75NTR (Jessen & Mirsky, 2019). In line with this, a gene expression study showed that nerve injury led to a partial recapitulation of the gene expression programs found in developing nerves. However, the same study also showed that a substantial subset of genes, including transcripts involved in cell structure and immune response, were specifically upregulated in injured nerves, distinct from developing or uninjured adult nerves (Bosse et al., 2006). More recently, repair Schwann cells were shown to also express a number of genes de novo, including Olig 1 and Shh, that are low/absent in immature Schwann cells during development (Arthur-Farraj et al., 2012; Jessen & Arthur-Farraj, 2019). These studies support the idea that repair Schwann cells are a distinct cell type in the Schwann cell lineage.

In addition, this becomes evident by the sharp contrast in the function of Schwann cells for myelin homeostasis. Whereas in uninjured nerves, Schwann cells are required for the generation and maintenance of myelin, in injured nerves, on the other hand, Schwann cells play an active role in the breakdown of myelin. It is estimated that Schwann cells are responsible for the breakdown of about 50% of total myelin during the initial days after injury, with the rest phagocytosed and degraded by resident and invading macrophages (Brosius Lutz et al., 2017; Hirata & Kawabuchi, 2002; Perry, Tsao, Fearn, & Brown, 1995). Schwann cells activate the process of myelinophagy, a form of selective autophagy, to digest their own myelin sheath both after nerve cut (Gomez-Sanchez et al., 2015; Jang et al., 2016) or nerve crush (Brosius Lutz et al., 2017). In this process, the myelin sheath is internalized and broken down into smaller cytoplasmic fragments before delivery to the lysosome for degradation (Jung et al., 2011). In addition, Schwann cells are also able to activate TAM receptor phagocytic pathways to engulf and break down smaller amounts of myelin (Brosius Lutz et al., 2017). This key role of Schwann cells and macrophages to clear myelin debris, which contains multiple nerve outgrowth inhibitors, together with the suppression of the myelin differentiation program as described above, are widely believed to enable robust axonal regeneration in the PNS (Jessen & Mirsky, 2019).

Nerve injury results in a massive influx of immune cells in the injured stumps and a recent single-cell RNA sequencing (scRNA-seq) study has revealed a vast complement of immune cells in injured nerves, including mast cells, B cells, T cells, natural killer cells, monocytes and macrophages (Carr et al., 2019). The infiltration of blood-derived macrophages into the injured nerves, which considerably outnumber resident macrophages (Mueller et al., 2003) is principally mediated by the secretion of the chemokine CCL2 by Schwann cells, which bind to the CCR2 receptor on monocytes. Genetic ablation of either Ccl2 or Ccr2 reduces the accumulation of macrophages in the injured distal nerve stump (Lindborg, Mack, & Zigmond, 2017; Niemi et al., 2013; Siebert, Sachse, Kuziel, Maeda, & Bruck, 2000). This increase in CCL2 expression in Schwann cells occurs downstream of the ERK-signalling pathway (Napoli et al., 2012) and possibly involves induction by cytokines, including TNF-α, LIF and IL-6, and engagement and activation of Toll-like receptors in Schwann cells (Boerboom et al., 2017; Zigmond & Echevarria, 2019). Macrophages play an important role in the breakdown of myelin debris together with the Schwann cells. Macrophages can also be recruited to sensory and sympathetic ganglia after injury in response to increased expression of CCL2 by axotomized neurons, where they participate in the activation of a transcriptional program in the neurons that promotes neurite outgrowth (Cattin & Lloyd, 2016; Kwon et al., 2013, 2015; Niemi, DeFrancesco-Lisowitz, Cregg, Howarth, & Zigmond, 2016; Niemi et al., 2013; Zigmond & Echevarria, 2019). In addition, prostaglandin D2 synthase, potentially via its expression in neurons, has recently been shown to modulate macrophage activity and accumulation after nerve injury (Forese et al., 2020).

Inflammatory cells (macrophages and neutrophils) together with fibroblasts are also key initial players in the formation of the nerve bridge, which is the new tissue formed between the proximal and distal stumps after complete nerve transection. Indeed, mice lacking CD11+ immune cells after acute nerve injury have been shown to suffer from poor axonal regeneration along with an impaired myelin clearance and compromised formation of blood vessels beyond the size of lesions (Barrette et al., 2008). In line with this, the hypoxic environment in the bridge was found to be selectively sensed by macrophages, which then secrete VEGF that promotes the formation of blood vessels across the bridge (Cattin et al., 2015; Cattin & Lloyd, 2016). Schwann cells then emerge from both stumps and migrate along the vasculature, carrying the regrowing axons from the proximal stump across the hostile environment of the bridge and into the distal stump (Cattin et al., 2015; Cattin & Lloyd, 2016; Stierli et al., 2018). In addition, they also upregulate the adhesion molecules NCAM and N-cadherin that have been linked to the enhanced migration-promoting properties of the Schwann cell surface (Arthur-Farraj et al., 2012; Napoli et al., 2012).

2.2 Axonal regeneration and remyelination

Axonal regrowth is initiated at the distal stump and slowly progresses around 1 mm/d, which corresponds to the rate of slow axonal transport (Höke & Brushart, 2010). While axonal regeneration occurs, in general, efficiently after axonotmesis, nerve transection (neurotmesis) often results in misdirected axonal regrowth and poor repair due to the missing continuity of the endoneurial tubes (Nguyen, Sanes, & Lichtman, 2002). In general, the prerequisite for axonal regeneration after acute nerve injury is the activation of genes that regulate cell survival and neurite outgrowth in neurons, referred to as the Regeneration-associated gene (RAG) program, as well as the local translation of mRNA and an increase in mitochondrial density to provide energy to the growing axons (Allodi, Udina, & Navarro, 2012; Bradke et al., 2012; Chen, Yu, & Strickland, 2007; Girouard et al., 2018). These multifaceted cell-intrinsic mechanisms include specific injury signals from the injury side to the neuronal soma (Allodi et al., 2012; He & Jin, 2016; Mahar & Cavalli, 2018; Terenzio, Schiavo, & Fainzilber, 2017).

The subsequent regrowth of axons starts with the formation of the growth cone, a highly mobile tip, which elongates following the endoneurial tubes and the Schwann cell Bands of Büngner (see below, Figure 1) in order to ultimately reinnervate the target organs (Allodi et al., 2012; Chen et al., 2007). While our understanding of the role of the multicellular microenvironment at the injury side as well as of the essential function of Schwann cells for axonal regeneration has substantially increased within the last decade (see below), our knowledge about signals from regrowing axons to their neighboring cells remains poor. Hence, to what extent axonal signals like growth factors, secreted molecules, neurotrophins, neurotransmitters, electrical activity or others impact Schwann cell alignment and redifferentiation during early axonal regeneration requires future studies.

In turn, regrowing axons are strongly dependent on the trophic support of Schwann cells for their survival, at the very least. The essential role of Schwann cells for successful regeneration of outgrowing axons to the target tissue has first been nicely demonstrated by experiments with acellular nerve grafts, in which axonal regeneration was severely impaired (Hall, 1986). Indeed, in the absence of Schwann cells, regenerating axons are misrouted, impairing the reinnervation of target organs (Rosenberg et al., 2014; Xiao et al., 2015). Since then, numerous studies have demonstrated that interference with Schwann cell function after nerve injury ultimately impairs axonal regeneration and functional recovery (for reviews see Jessen & Mirsky, 2016, 2019 and Jessen & Arthur-Farraj, 2019). While all of these studies illustrate the paramount importance of the overall Schwann cell integrity for nerve regeneration, our understanding of specific factors that signal in a glia to axon mode to promote axonal regeneration is still limited. The most direct form of glia to axon support during axonal regeneration constitutes the providence of physical guidance structures for regrowing axons. Here, Schwann cells adopt an elongated, bipolar shape in order to form columns from the injury site to the target tissue within the original basal lamina tube, which is essential for the robust and directed growth of regenerating axons. Lineage tracing experiments have demonstrated that both myelinating and non-myelinating Schwann cells generate the repair Schwann cells in injured nerves. Remarkably, this involves a threefold elongation of the Schwann cells that also often form extensive long processes that lie parallel to the main axis of the cell. This increases the cell-cell overlap of the Bungner cells to form the continuous and robust cellular regeneration lanes for the regrowing axons (Gomez-Sanchez et al., 2017). Indeed, mouse mutants in which the formations of Büngner Bands are compromised suffer from impaired axonal regeneration (Jessen & Mirsky, 2019; Scheib & Höke, 2013).

This is exemplified in mice with an ablation of c-Jun, an essential transcription factor for the acquisition of the Schwann cell repair phenotype (Arthur-Farraj et al., 2012). Disturbance of the repair phenotype of Schwann cells by ablation of c-Jun leads to a substantial increase in neuronal death as well as to defects in axonal regeneration, both after the injury to the facial nerves or sciatic nerves (Arthur-Farraj et al., 2012; Fontana et al., 2012).

These mice demonstrate abnormal regeneration tracks with flattened and irregular cellular profiles as well as an impaired expression of proteins that have been suggested to mediate axonal growth such as GDNF, artemin, BDNF, p75NTR and N-cadherin (Arthur-Farraj et al., 2012; Fontana et al., 2012). A severely impaired nerve repair also characterizes mice with a Schwann cell-specific ablation of the tumor suppressor Merlin, and Merlin-dependent signaling via the Hippo pathway and its effector YAP were found to contribute to the regulation of c-Jun and neurotrophin expression, including GDNF and Artemin (Mindos et al., 2017).

Notably, treatment with recombinant GDNF and Artemin partially rescued the axonal regeneration phenotype in Schwann cell–c-Jun deficient mice after acute nerve injury, suggesting that both factors contribute to axonal outgrowth, potentially via interaction with neuronally expressed Ret receptors (Fontana et al., 2012). In line with this, other studies also demonstrated a positive effect of treatment with different neurotrophins, including BDNF, GDNF, and Artemin on axonal regeneration (Boyd & Gordon, 2003; Gordon, 2009; Wang et al., 2014). In addition, p75NTR has been implicated in nerve regeneration and grafting experiments with p75NTR–deficient Schwann cells impairs neuronal survival after injury (Song, Zhou, Zhong, Wu, & Zhou, 2006; Tomita et al., 2007). However, the effects of different neurotrophins on nerve repair seem to also depend on the injury model and several neurotrophic factors are upregulated by both neurons and Schwann cells after injury.

Hence, the specific contribution of endogenous glial neurotrophins for axonal regeneration remains only partially understood (for reviews see Allodi et al., 2012; Boyd & Gordon, 2003; Gordon, 2009). Interestingly, neurotrophic factors have been shown to be differentially expressed by Schwann cells depending on their localization after nerve injury, suggesting that the Schwann cell phenotype may vary depending on their association with motor versus sensory fiber types (Brushart et al., 2013; Höke et al., 2006; Martini, Schachner, & Brushart, 1994). In line, Schwann cells support the pathfinding of axons and the glycosyltransferase Ih3 expression by Schwann cells has been shown to promote target selective regeneration of zebrafish axons after injury (Isaacman-Beck, Schneider, Franzini-Armstrong, & Granato, 2015).

Notably, the function of denervated Schwann cells to secrete growth factors in order to promote repair is not exclusive for the regeneration of nerves. In newts, which have the remarkable capacity to regenerate an entire limb, denervated Schwann cells secrete the growth factor newt anterior gradient protein that promotes the proliferation of the blastema, the population of stem cells formed at the cut site that generate the limb (Kumar & Brockes, 2012). For wound healing, after injury to the skin, denervated Schwann cells migrate into the wound bed, where they secrete factors that enhance TGFβ1 signalling to modulate the multicellular response and repair of the wound site (Johnston et al., 2013; Parfejevs et al., 2018). Finally, denervated Schwann cells can secrete growth factors, including platelet-derived growth factor (PDGF)-AA and oncostatin M that promote regeneration of the murine digit tip (Johnston et al., 2016).

Interestingly, Schwann cells have been demonstrated to secret exosomes after acute nerve injury and internalization of glial exosomes by DRG axons has been shown in vitro. Indeed, a treatment with Schwann cell-derived exosomes in injured nerves enhanced axonal regeneration in vivo, suggesting that a transfer of growth-supporting factors from Schwann cells to axons may contribute to nerve repair (Lopez-Leal & Court, 2016; Lopez-Verrilli, Picou, & Court, 2013).

Eventually, redifferentiated Schwann cells remyelinate newly outgrown axons (Figure 1). Notably, remyelinated segments are thinner and importantly, shorter, which impacts axonal function on the long term (see below). Axonal NRG1 expression is initially downregulated during Wallerian degeneration and axonal breakdown, but slowly returns back to normal expression levels within the time course of axonal regeneration, and most likely, in response to target reinnervation (Bermingham-McDonogh, Xu, Marchionni, & a, Scherer SS., 1997; Stassart et al., 2013). In line, axonal NRG1 signaling has been shown to play an important role in the redifferentiation and remyelination phase after injury (Fricker et al., 2011, 2013; Stassart et al., 2013). The transgenic overexpression of axonal NRG1 isoforms improves remyelination and myelin sheath thickness after nerve injury (Stassart et al., 2013). Interestingly, the overexpression of the soluble Neuregulin1 type I isoform in neurons reconstitutes the physiological myelin sheath thickness in regenerated nerves, although the same isoform does not promote primary Schwann cell myelination during development (Michailov et al., 2004; Stassart et al., 2013). Conversely, the ablation of axonal NRG1 (either by a SLICK-A-Cre or a CAG-Cre-ERTM line) leads to impaired remyelination with amyelinated and thinner myelinated regrown axons (Fricker et al., 2011). Surprisingly, this effect was transient, as mice with CAG-Cre-ERTM-mediated NRG1 ablation showed a similar remyelination extent to controls at three months post-injury (Fricker et al., 2013). Hence, these data would suggest that the mechanisms that drive developmental myelination and remyelination in the peripheral nervous system are fundamentally different. Of note, Schwann cell proliferation was not affected by NRG1 or ErbB receptor deletion after acute nerve injury in vivo (Atanasoski et al., 2006; Fricker et al., 2011).

Similar to NRG1, the ablation of Erbin, a protein that interacts with ErbB2 in Schwann cells, as well as the deletion of the NRG1 cleaving secretase BACE1 result in impaired remyelination after acute nerve injury (Hu et al., 2008; Hu, Hu, Dai, Trapp, & Yan, 2015; Liang et al., 2012). Interestingly, nerve graft experiments have suggested that both axonal and glial BACE1 contribute to remyelination (Hu et al., 2015). However, the clearance of myelin debris early after injury, as well as axonal regeneration, has also been reported to be negatively controlled by BACE1 (Farah et al., 2011). In line with this, neuronal BACE1 overexpression decreased axonal regeneration in vivo, suggesting that the timing and cellular localization of BACE1 may be important for different aspects of nerve repair (Farah et al., 2011; Tallon, Rockenstein, Masliah, & Farah, 2017). Hence, further studies are needed in order to better understand the role of BACE1 in nerve regeneration (Pellegatta & Taveggia, 2019).

Of note, axons are not the only source of NRG1 in injured nerves. Schwann cells themselves also induce a de novo expression of soluble NRG1 type I (both the alpha and beta isoform) within one day after nerve injury (Hu et al., 2015; Ronchi et al., 2013; Stassart et al., 2013 and own unpublished observation). Indeed, Wallerian degeneration and axonal breakdown transiently preclude transdifferentiating Schwann cells from access to axonal signaling cues, which triggers denervated Schwann cells to temporarily express NRG1 as an autoparacrine signal that supports Schwann cell redifferentiation and remyelination (Stassart et al., 2013). In line with this, while the ablation of Schwann cell-derived NRG1 did not affect developmental myelination, remyelination and functional recovery are impaired in respective mutant mice lacking Schwann cell Neuregulin-1 after acute nerve injury (Stassart et al., 2013). Hence, Schwann cell redifferentiation and remyelination may be less dependent on axonal signaling cues compared to the equivalent developmental processes.

3 CHRONIC NERVE INJURY

Chronic nerve injury describes any condition in which the regeneration process and subsequent functional recovery of acutely injured nerves are delayed or eventually fail. A prolonged chronic injury constitutes a common and severe problem in peripheral nerve injury in humans, as long peripheral nerves, such as the sciatic nerve, are specifically prone to chronic denervation (Höke, 2006; Höke & Brushart, 2010; Terenghi, Calder, Birch, & Hall, 1998). Indeed, it is generally accepted that the time to reinnervation represents the most important determinant of successful nerve repair, which is largely determined by the rate of axonal regrowth during regeneration that is around 1mm/d (Höke & Brushart, 2010).

What determines the regeneration failure in long peripheral nerves with time? Importantly, neuronal cell bodies of adult axotomized nerves survive for long periods of time (Carlson, Lais, & Dyck, 1979; Gordon, Gillespie, Orozco, & Davis, 1991) and cross suture experiments indicate that although the number of regenerating axons strongly decreases with time after injury, functional recovery of muscle fibers can be achieved if a chronically axotomized nerve is joined to a freshly denervated distal nerve part (Fu & Gordon, 1995; Holmes & Young, 1943). Conversely, cross suturing of a freshly axotomized nerve into a long-term denervated distal nerve stump, however, caused poor functional recovery of respective muscle fibers, suggesting that the distal nerve environment, and not intrinsic neuronal/axonal properties may determine peripheral nerve regeneration outcome (Fu & Gordon, 1995; Sulaiman & Gordon, 2000).

Here, Schwann cell numbers as well as the molecular phenotype of these cells in chronically denervated nerves may constitute two key factors impacting nerve repair in the long term. As discussed above, Schwann cell proliferation represents an integral part of the early repair responses after nerve injury with an approximately 2 to 3-fold increase in Schwann cell numbers in the first week of post-injury (Jessen & Mirsky, 2019). Based on different studies, Schwann cell numbers subsequently strongly drop around 2 months after chronic denervation, though they still remain higher compared to cell numbers in non-lesioned controls (Jessen & Mirsky, 2019; Jonsson et al., 2013; Salonen, Aho, Röyttä, & Peltonen, 1988; Siironen, Collan, & Röyttä, 1994). Whether a threshold number of Schwann cells is necessary for successful regeneration and hence, whether the observed reduction in Schwann cell numbers in chronic denervation may contribute to impaired regeneration, remains only partially understood. Studies of cyclin D1 deficient mice that are characterized by a reduced Schwann cell proliferation after nerve injury suggest that absolute Schwann cell numbers may be of limited importance for nerve repair (Atanasoski, Shumas, Dickson, Scherer, & Suter, 2001; Kim et al., 2000; Serhan, Yacoubian, & Yang, 2008). However, as Schwann cell numbers in cyclin D1-/- mutant mice return to wild-type numbers within two weeks after injury (Yang et al., 2008), the contribution of cell numbers for the following steps of nerve repair remains to be determined.

Notably, a more recent study has shown that the genetic inactivation of the transcription factor STAT3 in Schwann cells, which is normally activated after acute nerve injury, causes a decrease of autocrine Schwann cell survival loops, along with a strong reduction of Schwann cell numbers after injury, and impairs nerve regeneration (Benito et al., 2017). However, STAT3 deletion also downregulates the expression of Schwann cell repair markers in chronically injured nerves, leaving open the possibility that Schwann cell death may be secondary to a loss of the Schwann cell repair phenotype (Benito et al., 2017).

Indeed, the drop of Schwann cell numbers in chronic denervation presumably arises as a consequence of a gradual fading of the Schwann cell repair phenotype with increasing time after injury (Jessen & Mirsky, 2019), and molecular correlates of

留言 (0)

沒有登入
gif