The Contribution of Innervation to Tissue Repair and Regeneration

Abstract

Animals such as amphibians have an incredible capacity for regeneration with some being able to regrow their tail or appendages. Although some mammalian tissues like the skin and bones can repair following injury, there are only a few examples of true multilineage regeneration, including the distal portion of the digit tip. In both amphibians and mammals, however, to achieve successful repair or regeneration, it is now appreciated that intact nerve innervation is a necessity. Here, we review the current state of literature and discuss recent advances that identify axon-derived signals, Schwann cells, and nerve-derived mesenchymal cells as direct and indirect supporters of adult tissue homeostasis and repair. We posit that understanding how nerves positively influence repair and regeneration could lead to targeted regenerative medicine strategies to enhance tissue repair in humans.

INTRODUCTION TO THE PERIPHERAL NERVOUS SYSTEM

The peripheral nervous system (PNS) is one of the most widespread and complex structures in the mammalian body, being omnipresent in all tissues and organs. The PNS includes a variety of different structures, including receptor cells that sense external stimuli as diverse as heat, scents, and oxygenation levels, peripheral neurons that are located within larger ganglia or are embedded locally in tissues, and the peripheral nerves that link peripheral neurons with the tissues they innervate. The PNS has been extensively studied by neuroscientists, because it is the main conduit for information traveling to and from the central nervous system (CNS) and the body. In this review, we will not focus on these well-characterized functions, but will instead focus on a more recently appreciated, noncanonical role for peripheral nerves in the regulation of tissue repair and regeneration. In the remainder of the review we will use the term regeneration to refer to cases in which complex anatomic structures are restored so that they are morphologically and functionally similar to the structure before injury. We will use the term repair to refer in a more general way to mechanisms that allow restoration of at least some level of functionality following tissue damage.

Peripheral nerves vary considerably in size and are frequently defined by the type of axons that they carry and, thus, their functional roles. For example, motor nerves carry the axons of motor neurons that regulate muscle contraction, sensory nerves carry information from sensory receptors back to the CNS, whereas large nerves such as the well-studied sciatic nerve are considered to be mixed nerves because they carry axons of multiple different types of neurons. Nonetheless, despite these functional differences, peripheral nerves all have a similar underlying tissue structure. The fundamental unit of peripheral nerves are the fascicles (Fig. 1) that contain neuronal axons and their associated glial cells, Schwann cells, which may or may not be myelinating, depending on the size and type of their associated axons. Notably, fascicles also contain a third, much more poorly characterized cell type, endoneurial fibroblasts. These mesenchymal cells are interspersed among and closely associated with Schwann cells and axons, and we are only now starting to appreciate their important functional roles (see below). Each individual fascicle is surrounded by a layer of perineurial cells, a second type of mesenchymal cell that forms an epithelia-like barrier. The fascicular bundles are then surrounded by a relatively loose mesenchymal cell–derived connective tissue stroma that contains embedded vasculature, immune cells, and, in large nerves, adipocytes. This final stromal compartment is called the epineurium and is conceptually similar to the connective tissue sheaths that wrap most organs in the body. Historically, with the exception of Schwann cells, these nerve-associated cell types were relatively poorly understood, but a recent series of single-cell transcriptomic studies have provided many new insights into their identity under both homeostatic and injury conditions (Carr et al. 2019; Toma et al. 2020; Wolbert et al. 2020; Gerber et al. 2021).

Figure 1.

Peripheral nerve anatomy. Peripheral nerves are composed of three connective tissue layers termed the epineurium, perineurium, and endoneurium. Nerve axons, Schwann cells (myelinating and nonmyelinating), and endoneurial fibroblasts reside within the endoneurium of nerve fascicles, which are ensheathed by a connective tissue membrane and perineurial fibroblast-like cells. Immune and vascular-related cells are embedded in connective tissue, which is interspersed between nerve fascicles that are collectively ensheathed by a connective tissue epineurial stroma.

The most extensively studied nerve cells are Schwann cells, which play essential roles in regulating axon physiology, maintaining axonal health, and promoting axon regeneration following injury (for an in-depth review of this topic, see Jessen et al. 2015). Developmentally, Schwann cells derive from the neural crest, a population of multipotent and migratory embryonic cells with incredible cellular flexibility that generate diverse cell types of both neural and mesenchymal lineages (Le Douarin 1986; Kaucká and Adameyko 2014). These embryonic neural crest stem cells generate downstream progenitors that are called Schwann cell precursors (Jessen and Mirsky 2019a), because they show many Schwann cell–like properties. Nonetheless, lineage-tracing studies have shown that, at least during embryogenesis, these Schwann cell precursors give rise to multiple cellular lineages, including, for example, mature Schwann cells, peripheral neurons, and peripheral tissue melanocytes (Furlan and Adameyko 2018). Perhaps most surprisingly, embryonic Schwann cell precursors also generate endoneurial fibroblasts (Joseph et al. 2004), but not the other nerve mesenchymal populations, which are instead thought to arise from the mesoderm.

The peripheral nerve displays a remarkable ability for repair and regeneration (Cattin and Lloyd 2016; Jessen and Mirsky 2019b), a capacity that may have arisen as a consequence of its relatively vulnerable anatomy and the necessity for ensuring that the neurons whose axons it carries persist and function throughout our lifetimes. This pro-regenerative ability is thought to be a result of close interactions between Schwann and mesenchymal cells that rapidly repair nerve damage and then produce an environment conducive to axonal regeneration (Parrinello et al. 2010; Cattin et al. 2015). Many studies asking about this pro-regenerative environment support the concept that it involves trophic factors made by Schwann cells (Nave 2010; Jessen and Mirsky 2016; Weiss et al. 2016) and perhaps even nerve mesenchymal cells (Toma et al. 2020). The nerve's capacity for regeneration, ability to provide trophic support, and widespread distribution in all tissues and organs have led to the idea that peripheral nerves may also play a much broader role in tissue repair and regeneration, an idea with increasing support as discussed in this review.

NERVE-DEPENDENT REGENERATION AND THE FUNCTION OF SCHWANN CELLS

One major goal of regenerative medicine is the restoration of tissue structure and function following injury. To develop strategies for accomplishing this goal, researchers often turn to studies examining why species such as salamanders and newts can perform spectacular feats of regeneration (restoration of functional complex anatomical structures), whereas others, such as mammals, largely heal by repair where in many cases they form a tissue substitute such as a scar (Tanaka 2016). One increasingly clear conclusion of this work is that peripheral nerves are an important source of trophic factors and even cellular components during tissue regeneration. This conclusion largely derives from seminal work performed in amphibians (see below), but similar conclusions have been reached studying other nonmammalian model systems such as the sensory barbels of catfish, which are supported by a cartilaginous core and nerves and starfish arm regeneration, which is contingent on an oral nerve ring that projects to the radial arm nerves (for review, see Kumar and Brockes 2012).

As stated, amphibians represent the best-documented example of a requirement for nerve innervation in tissue regeneration. In salamanders, limb regeneration has been thought of as a sequel to embryonic limb development and growth (Kumar and Brockes 2012) because many of the same cues and cell types are involved. However, one major divergence is that nerves are not required for the outgrowth of the embryonic limb (Brockes and Kumar 2008) but they are essential for regeneration. Following amputation, salamander limb regeneration is initiated by the formation of a specialized epithelium with the subsequent establishment and expansion of a regenerative mass of cells known as a blastema that ultimately gives rise to the regenerated limb mesenchyme (Tanaka 2016). Initial work showed that when nerves were removed by axotomy before limb amputation, regeneration ultimately failed because of poor expansion of the blastema (Butler and Schotté 1941). Although these early studies showed that nerves were essential, it was not clear whether this reflected a requirement for the innervating axons or for resident nerve cell types. This question was answered when Kumar and Brockes (Kumar et al. 2007; Kumar and Brockes 2012) showed that nerve-derived Schwann cells provided essential trophic support for the expanding blastema, and identified a ligand, newt anterior gradient (nAG) protein that was essential for this activity (Kumar et al. 2007). These studies were instrumental in establishing that Schwann cells provided trophic support not only for regenerating axons, but also for regenerating peripheral tissues. However, subsequent work showed that nAG was not apparently expressed in mammals and that its closest homologs (AGR2, AGR3) displayed very restricted expression (Komiya et al. 1999). More recent work showed that ARG2 is up-regulated in cancers of the breast, lung, ovarian, and prostate (Patel et al. 2013), but it has not been implicated in mammalian tissue regeneration (Ivanova et al. 2013). Nonetheless, these studies were some of the first to describe a role for Schwann cells in an “extraneural” capacity, and set the stage for more recent work examining Schwann cell function during the repair and regeneration of mammalian tissues, as described below.

NERVES ARE NECESSARY FOR MAMMALIAN DIGIT TIP REGENERATION

One notable exception to the general rule that mammals do not regenerate is the tip of the digit/finger. In rodents, monkeys, and humans (Shieh and Cheng 2015), removal of the distal one-third of the terminal phalangeal element results in complete regeneration and proper patterning of the lost structure, as long as the proximal nail components are preserved. Like amphibian limb regeneration, digit tip regeneration involves formation of a blastema, which is largely comprised (85%) of proliferating platelet-derived growth factor receptor α (PDGFRα)-positive mesenchymal cells (Johnston et al. 2016; Carr et al. 2019; Storer et al. 2020). Immediately following amputation of the adult digit tip within the regeneration-competent region, there is an initial retraction of the injured epidermis, immune cell infiltration, local bone degradation, and closure of the wound epidermis through the newly eroded bone. These early events are followed by formation of the blastema, which is initiated at about 7 d postamputation in adult mice. The blastema expands to a maximal size by 10–14 d, and starts regenerating the lost digit tip tissues at the same time. Regenerative growth is finished by ∼1 mo postamputation, and this is followed by a 4 wk remodeling period that culminates in complete digit tip restoration (for reviews, see Storer and Miller 2020, 2021).

Because the key essential event in digit tip regeneration is formation of a blastema, recent studies have focused on understanding the identity and origin of the blastema cells (Lehoczky et al. 2011; Rinkevich et al. 2011; Simkin et al. 2013; Shieh and Cheng 2015; Johnston et al. 2016; Carr et al. 2019; Storer et al. 2020). Although the blastema was historically thought to be a homogeneous mixture of multipotent precursor cells, this more recent work indicates that it is instead a heterogeneous mixture of locally derived cells that each regenerate cells of their own lineages. The most abundant of these diverse cells are PDGFRα-expressing mesenchymal precursors that derive from local mesenchymal tissues such as the bone and dermis and then regenerate only the lost mesenchymal tissues. Intriguingly, single-cell transcriptomic approaches show that when these tissue-derived mesenchymal cells contribute to the blastema, they lose their tissue-specific gene expression and acquire a unique blastema transcriptional state that includes genes involved in both development and adult tissue injury (Storer et al. 2020). These mesenchymal blastema cells then contribute in an origin-independent manner to mesenchymal regeneration. For example, bone lineage cells ultimately contribute to regeneration of both the bone and dermis. Together, these studies support a model in which regenerative digit tip amputations somehow recruit cells from local mesenchymal tissues, induce them to undergo dedifferentiation to a unique blastemal precursor state, and then these blastema cells regenerate all of the local mesenchymal tissues in an origin-independent fashion.

What then do peripheral nerves have to do with this? Reinnervation of the regenerating digit tip occurs around 10–14 d postamputation, concomitant with the initial formation and expansion of the blastema (Johnston et al. 2016). Like amphibian limbs, mammalian digit tip regeneration is contingent on nerve innervation. Surgical denervation before digit tip amputation delays blastema formation and growth (Takeo et al. 2013), thereby impairing regeneration (Mohammad and Neufeld 2000; Johnston et al. 2016) and elicits notable patterning defects (Rinkevich et al. 2014). A number of recent reports indicate that nerves mediate these pro-regenerative actions both by Schwann cell–dependent trophic mechanisms and by direct contribution of cells for mesenchymal tissue regeneration (Johnston et al. 2016; Carr et al. 2019), both of which will be discussed in detail below.

SCHWANN CELLS PROVIDE TROPHIC SUPPORT FOR BLASTEMA EXPANSION AND DIGIT TIP REGENERATION

As discussed above, mammalian Schwann cells are well known for their pro-regenerative actions within a nervous system context. Following injury, peripheral nerves undergo a process known as Wallarian degeneration, which involves a cascade of cellular events aimed at clearing axon and myelin debris (Cattin and Lloyd 2016) and promoting axon regrowth. During this process, Schwann cells increase their expression of c-jun and the stem cell gene Sox2, transcriptional changes that play a role in reprogramming them into a migratory, proliferative state that is necessary to support myelin clearance and axon regeneration and guidance (Jessen and Mirsky 2019b). These Sox2-positive dedifferentiated Schwann cells are well known for their robust expression of trophic factors that are essential for axonal regeneration (Jessen et al. 2015; Cattin and Lloyd 2016; Jessen and Mirsky 2019b; Toma et al. 2020). Indeed, classic studies performed by Aguayo and others (Aguayo et al. 1977, 1981; So and Aguayo 1985; Bray et al. 1987) showed that transplantation of peripheral nerves or their resident Schwann cells was sufficient to promote the regeneration of CNS axons that normally do not regenerate.

What happens to Schwann cells during digit tip regeneration? Amputation causes local nerve damage and, consistent with this, Sox2 is induced in Schwann cells associated with the injured nerves, indicative of their dedifferentiation (Johnston et al. 2013, 2016). These dedifferentiated Schwann cells then detach from the degenerating nerves and migrate toward and around the newly formed blastema. Four lines of evidence indicate that these nerve-derived Schwann cells are essential for normal digit tip regeneration. First, when the sciatic nerve is transected 10–14 d before amputation, this leads to a loss of both axons and Schwann cells from the digit tip, and ultimately results in impaired blastema proliferation and expansion, and decreased nail and bone regeneration. Second, a similar deficit in regeneration is observed when dedifferentiated Schwann cells are specifically ablated using a targeted diphtheria toxin subunit A (DTA)-mediated killing strategy. Third, when Sox2 is inducibly knocked out before digit tip amputation, this leads to a decreased number of dedifferentiated Schwann cells within the blastemal region, and an abrogation of bone and nail regeneration. Finally, when cultured dedifferentiated Schwann cells were transplanted into the denervated, amputated digit tip, they were able to rescue the denervation-induced deficits in blastema expansion and the extent of nail and bone regeneration back to normal levels (Johnston et al. 2016). Intriguingly, these transplanted Schwann cells could not rescue the observed patterning deficits, arguing that their precise localization is important.

Together, these results make a compelling argument that, as during amphibian limb regeneration, nerve-derived Schwann cells play a key role in promoting blastema expansion and ultimately tissue regeneration. How do they do this? One potential explanation is that they might directly contribute mesenchymal cells for tissue regeneration as Schwann cell precursors do during cranial mesenchymal embryogenesis (Xie et al. 2019). However, lineage tracing following digit tip amputation found no evidence that adult dedifferentiated Schwann cells contributed to anything other than more Schwann cells (Johnston et al. 2016). Moreover, recent single-cell RNA transcriptome analyses indicate that injured adult nerve dedifferentiated Schwann cells are transcriptionally distinct from their developing counterparts (Toma et al. 2020).

A second explanation is that dedifferentiated Schwann cells promote digit tip regeneration by secreting ligands, as they do during nervous system regeneration (Fig. 2). To test this idea, Johnston et al. (2016) used transcriptomic and proteomic data to model potential paracrine interactions and, in so doing, identified platelet-derived growth factor-AA (PDGF-AA) and oncostatin M (OSM) as ligands that were secreted by dedifferentiated Schwann cells and that had receptors on blastema cells. Importantly, both OSM and PDGF-AA are up-regulated in dedifferentiated Schwann cells in vivo following tip digit amputation and exogenous OSM and PDGF-AA were able to rescue the denervation-induced deficits in digit tip regeneration (Johnston et al. 2016). Intriguingly, PDGF-AA is known to promote growth of other mesenchymal tissues, including skin, bones (Andrae et al. 2008; Donovan et al. 2013), and hair follicles (Tomita et al. 2006) and although little is known about OSM as a Schwann cell ligand, this IL-6 family member has been implicated in bone mineralization (Walker et al. 2010) and mesenchymal precursor to osteoblast differentiation (Sims and Quinn 2014). Thus, dedifferentiated Schwann cells might broadly regulate mesenchymal tissue repair via these two ligands.

Figure 2.

Schematic of the nerve dependency of digit tip regeneration. (A) Anatomy of the uninjured mouse digit tip demonstrating the plane of amputation, which results in successful regeneration. (B) Mouse digit tip following amputation, which has formed a regenerative blastema with associated cell types. Amputation of the digit tip results in peripheral nerve damage and the subsequent dedifferentiation of Schwann cells. These Sox2-positive Schwann cells then migrate into the blastema and secrete growth factors like PDGF-AA and OSM that promote blastema cell proliferation. Coincident with this, nerve-derived mesenchymal cells migrate from damaged nerves into the blastema where they have the potential to differentiate into mesenchymal tissue types such as bone and dermal cells. Collectively, these processes provide cellular components and paracrine growth factors that facilitate digit tip regeneration.

The aforementioned work focuses on one of the few mammalian systems in which there is multitissue regeneration, but most mammalian tissues do not regenerate. Nonetheless, there is a broad literature demonstrating that when mammalian tissues are denervated, this results in poor tissue repair and wound healing. For example, injury to denervated skeletal muscle elicits fibrosis and poor muscle repair (Madaro et al. 2018) and reducing the activity of sensory nerve fibers in fractured bones delays ossification of the fracture callus (Li et al. 2019). Additionally, loss of local innervation within the skin increases the risk of diabetic foot ulceration by 8- to 18-fold (Nowak et al. 2021). These findings suggest that Schwann cells might play a broad and generalizable role in promoting tissue repair and, if so, raise the possibility that this potential could be harnessed therapeutically. These important questions are being addressed in a variety of systems, but the most compelling studies examine skin and bone repair, both of which will be discussed below.

SKIN WOUND HEALING: IMPORTANT ROLES FOR INNERVATION AND SCHWANN CELLS

When the skin is injured, this triggers a cascade of distinct, but overlapping responses involving many cell types and progenitors including immune cells, dermal fibroblasts, keratinocytes, and nerves, among others. Briefly, the immediate response is hemostasis, which is typified by blood vessel constriction and activation of platelets to form a fibrin clot. This is followed by high immune cell activity, which stimulates the proliferation and migration of epithelial cells, fibroblasts, and endothelial cells that ultimately seal the wound, create granulation tissue, and facilitate angiogenesis (Gurtner et al. 2008). Eventually, the healed area is remodeled by aligning and strengthening collagen fibers as the tissue matures (Gurtner et al. 2008).

What do innervation and/or Schwann cells have to do with these responses? Both human and rodent skin is richly innervated with a dense network of sympathetic axons and sensory nerve afferents. As a consequence, damage to the skin elicits not only the previously described local nerve degeneration and Schwann cell activation, but also a robust increase in the secretion of many axon-derived neuropeptides (Fig. 3; da Silva et al. 2010). However, there are conflicting reports regarding the necessity of these neuron-derived neuropeptides during skin wound healing (Buckley et al. 2012; Rinkevich et al. 2014). Some studies provide evidence that these neuropeptides are important. For example, mice lacking the neuropeptide Y receptor (NPY-2Ra) (Ekstrand et al. 2003) or calcitonin gene-related peptide (CGRP) (Toda et al. 2008) show a delay in wound healing kinetics, and topical treatment of diabetic mice with the neuropeptides neurotensin (NT) or substance P (SP) accelerated wound closure by reducing inflammation and inducing fibroblast proliferation and collagen production (Moura et al. 2014; Kant et al. 2015; Um et al. 2017). Conversely, Wallengren et al. (1999) reported that there were no deficits in wound healing in wild-type mice following skin denervation or the partial destruction of peripheral axon terminals with capsaicin.

Figure 3.

Schematic of the nerve dependency of skin repair. Wounding of skin damages cutaneous axons and elicits dedifferentiation of Schwann cells. Damaged nerve axons provide trophic support for wound healing by secreting neuropeptides such as neuropeptide Y (NPY), calcitonin gene-related peptide (CGRP), neurotensin, and substance P. Sox2-positive Schwann cells migrate from damaged nerves into the healing dermis and promote the proliferation of dermal cells through the secretion of oncostatin M (OSM) and platelet-derived growth factor-AA (PDGF-AA). Platelet-derived growth factor receptor α (PDGFRα)-positive nerve mesenchymal cells provide cellular components for the healing skin by migrating from damaged cutaneous nerves and give rise to dermal cells.

However, although a role for axon-derived neuropeptides in skin healing is as-yet inconclusive, recent evidence suggests a more definitive function for Schwann cells in wound repair (Fig. 3; Johnston et al. 2016; Parfejevs et al. 2018). Under homeostatic conditions, Schwann cells are located in skin nerves or are associated with the fine axon terminals that are located directly within the skin (Biernaskie et al. 2009; Johnston et al. 2013). Following damage, the local axons and nerves are injured, and Schwann cells become activated and reprogrammed. These Sox2-positive dedifferentiated Schwann cells then migrate into the healing dermis where they are located immediately adjacent to the dermal mesenchymal cells (Johnston et al. 2016; Parfejevs et al. 2018). Several lines of evidence indicate that, as seen during digit tip regeneration, these dedifferentiated Schwann cells then act in a paracrine fashion to promote dermal repair. First, conditional ablation of Sox2-positive Schwann cells compromised wound repair by reducing dermal and epidermal cell proliferation (Johnston et al. 2016). Second, conditional deletion of the essential Schwann cell gene Sox10 before wounding decreased the number of dedifferentiated Schwann cells and impaired wound closure, likely caused by reduced myofibroblast numbers (Parfejevs et al. 2018). Third, when Sox2 was conditionally deleted, dedifferentiated Schwann cell numbers were reduced and wound healing was impaired (Johnston et al. 2013). Finally, culture studies showed that Schwann cells secrete ligands that promote proliferation and self-renewal of dermal mesenchymal precursors in culture (skin-derived precursors [SKPs]) (Johnston et al. 2016). Together, these studies support a model in which Schwann cell–derived ligands promote dermal precursor proliferation and thus enhance repair. However, these studies did not specifically identify the relevant ligands, and Schwann cells are thought to secrete many different ligands that might mediate these effects, including PDGF-AA (Toma et al. 2020). It will thus be interesting to determine which of these Schwann cell–derived factors are relevant in vivo, and to ask whether Schwann cells or their secreted ligands can promote wound healing in conditions in which this is impaired such as in diabetes.

It should be noted that under most circumstances, following wounding, the skin heals through reparative scar formation. However, under certain circumstances, some mammals such as mice and rabbits are able to regenerate functional, anatomically correct skin following injury (Iismaa et al. 2018). One of the best described examples of this is the skin regeneration elicited by significant skin trauma termed “wound-induced hair follicle neogenesis (WIHN)” (Taylor 1949). In this model, implementation of large 1.5 cm wounds to the back skin of mice can result in de novo hair follicles capable of cycling in addition to local adipocyte regeneration (Ito et al. 2007; Plikus et al. 2017). Perhaps more surprisingly are the existence of so-called “super-regenerator” mice, which are capable of robust and speedy regeneration following injury. For example, the African spiny mouse (Acomys) can heal excisional and burn-induced wounds at an accelerated rate compared with its Mus counterparts with the resulting tissue forming without a scar (Seifert et al. 2012; Maden and Varholick 2020). Within the healed skin, hair follicles, complete with sebaceous glands, are present. These incredible animals can do this repeatedly as in response to predation; their skin is easily torn to allow escape, after which regeneration ensues. Our current understanding of the molecular cues responsible for WIHN (Abbasi et al. 2020) and regeneration in Acomys (Gaire et al. 2021) or similar animals is only recently coming to light and present an opportunity to further interrogate the involvement of innervation in these processes.

SCHWANN CELLS ARE ALSO IMPORTANT FOR BONE REPAIR

A second system in which nerves are thought to be important is bone repair. Bones and their associated bone marrow are innervated by sensory, sympathetic, and parasympathetic axons, and there is accumulating evidence that signals deriving from both axons and Schwann cells are involved in multiple aspects of bone homeostasis, development, and repair (García-Castellano et al. 2000; Tomlinson et al. 2020). With regard to axons, several studies indicate that neuron-derived neurotransmitters regulate the biology of osteoblasts, osteoclasts, and mesenchymal stromal progenitors (Costa et al. 2011; Gharibi et al. 2011; Negishi-Koga et al. 2011). As one example, Semaphorin 3a, which is known to be essential for normal bone mass, is secreted by sensory nerves that innervate the bone. Notably, neuron-specific deletion of Semaphorin 3a caused perturbed bone formation and a reduction in bone mass (Fukuda et al. 2013). As a second example, catecholamine release by sympathetic neurons has been shown to uncouple bone formation and resorption (Ma et al. 2011). Indeed, secretion of norepinephrine stimulates β2-adrenergic receptors (β2AR) on osteoblasts to suppress their proliferation while at the same time enhancing osteoclastogenesis (Fu et al. 2005). Thus, axon-derived signals have the potential to regulate bone development and homeostasis, even when there is no bone damage.

Increasing evidence also highlights a role for Schwann cells in the regulation of multiple aspects of bone homeostasis, including, for example, regulation of the hematopoietic stem cell niche (Yamazaki et al. 2011; Isern et al. 2014). However, of most relevance for this review, there is accumulating evidence that Schwann cells might also promote bone repair (Fig. 2). Of particular note, Longaker and colleagues (Jones et al. 2019) recently showed that jaw denervation led to local Schwann cell depletion coincident with impairments in endochondral ossification and cartilage formation after mandible injury. Intriguingly, they attributed these deficits to poor expansion and activity of endogenous skeletal stem cells (SSCs) in the absence of Schwann cells (Jones et al. 2019). As seen in the digit tip regeneration studies, transplanted Schwann cells partially rescued the denervation-induced deficits in bone repair, as did the Schwann cell–derived factors PDGF-AA and OSM. Together with the digit tip regeneration and skin wound healing studies, these results suggest that following tissue damage, local Schwann cells secrete ligands that play a general role in promoting mesenchymal tissue repair and raise the possibility that Schwann cells might also secrete injury-induced ligands that enhance repair of other, nonmesenchymal tissues.

The aforementioned studies indicate that Schwann cells predominantly mediate their reparative effects via paracrine mechanisms. However, one important exception to this generalization involves the rodent tooth. In adult mice, incisor teeth grow continually throughout the life of the animal. There is general consensus that this involves a population of adult mesenchymal precursor cells that regenerate the tooth mesenchyme, although there is considerable discussion as to the origin and location of these cells. What is clear is that incisor regeneration requires innervation because removal of nerves results in aberrant tooth growth (Zhao et al. 2014). This deficit has been attributed to two distinct mechanisms. One study attributed it to nerve-dependent sonic hedgehog (Shh) signaling in mesenchymal precursors (Zhao et al. 2014). A second study by Igor Adameyko's group (Kaukua et al. 2014) instead attributed it to a novel Schwann cell mechanism. Specifically, they showed that a significant portion of tooth mesenchymal precursors arise developmentally from embryonic Schwann cell precursors. Surprisingly, they also found that the Schwann cell lineage directly contributes tooth mesenchymal precursors in adulthood (Kaukua et al. 2014), a finding that is quite different from the lineage tracing conclusions in the digit tip and skin studies (Johnston et al. 2013, 2016; Carr et al. 2019). Intriguingly in this regard, the tooth mesenchymal cells are neural crest derived, whereas the digit tip and skin mesenchymal cells are mesodermally derived, and these different results might thus reflect fundamental differences between regeneration and/or repair of neural crest–derived and mesodermally derived tissues.

NERVE-DERIVED MESENCHYMAL CELLS DIRECTLY CONTRIBUTE TO DIGIT TIP REGENERATION AND SKIN REPAIR

Together, these studies strongly support the conclusion that one major way nerves regulate tissue homeostasis and repair is via axon and Schwann cell–mediated paracrine mechanisms. However, mesenchymal cells constitute a major cellular component of all peripheral nerves, raising the question of whether they too play an important role. Earlier work on nerve mesenchymal cells largely focused on extracellular matrix deposition and turnover, and a potential barrier function for the perineurial cells (Mizisin and Weerasuriya 2011; Richard et al. 2012). However, several recent studies by Lloyd and colleagues showed that nerve-derived mesenchymal cells play an essential role in nerve repair following peripheral nerve injury (Parrinello et al. 2010; Cattin and Lloyd 2016) suggesting that they might also be important for tissue repair. This possibility was directly addressed in a recent paper by Carr et al. (2019) who focused on the regenerating digit tip (Fig. 2). Specifically, the investigators performed single-cell transcriptional profiling to identify the three different nerve mesenchymal cell populations, epineurial, endoneurial, and perineurial, and to characterize their response to sciatic nerve injury. Intriguingly, this analysis showed that the neural crest–derived endoneurial cells displayed a transcriptional profile reminiscent of mesenchymal precursor cells, suggesting that they might be able to directly contribute mesenchymal cells to regenerating mesenchymal tissues. Several lines of evidence indicated that this was indeed the case. First, when nerve mesenchymal cells were cultured under appropriate conditions, they differentiated into bone cells and adipocytes. Second, to see whether these cells could incorporate into regenerating bone, the investigators transplanted an isolated genetically tagged nerve segment adjacent to a bone injury. They found that nerve-derived mesenchymal cells migrated from the transplanted nerve into the bone repair callus in which they expressed bone genes such as Osterix. Third, they asked whether nerve-derived mesenchymal cells could do this endogenously, taking advantage of the neural crest origin of the endoneurial fibroblasts. Surprisingly, ∼20% of the digit tip mesenchymal blastema cells arose from nerve-derived mesenchymal cells and these cells ultimately contributed differentiated osteocytes and fibroblasts to the regenerated bone and dermis, which persisted for at least 84 d postamputation. Finally, the investigators showed that denervation before digit tip amputation abolished the contribution of nerve-derived mesenchymal cells to bone and dermis formation, as predicted.

These studies raised the possibility that peripheral nerves provide a general source of mesenchymal precursors for rebuilding injured mesenchymal tissues. Carr et al. (2019) examined skin repair to test this idea and showed that transplanted peripheral nerves could indeed contribute fibroblasts for dermal repair. They also showed that following skin wounding, two populations of cells emerged from the endogenous injured nerves, nerve-derived mesenchymal cells, and dedifferentiated Schwann cells. Both populations were found in and around the healing dermis (Carr et al. 2019). Thus, the peripheral nerve contributes mesenchymal cells for both repair and regeneration in several distinct injury paradigms (Figs. 2 and 3). Notably, recent work shows that these nerve mesenchymal cells express growth factors that can promote axonal growth (Toma et al. 2020), raising the possibility that, like Schwann cells, they too provide paracrine support for injured tissues. It will be important in the future to determine the relative importance of these different nerve-dependent mechanisms, particularly as we move forward to develop therapeutic strategies to promote tissue repair.

CONCLUDING REMARKS

It has become increasingly clear that nerves and nerve-associated cells function in a much broader capacity than once appreciated, and that they play a general role in promoting tissue repair and regeneration. One way that they do so is via Schwann cell–dependent growth factors, something that is important during nerve and digit tip regeneration, and skin and bone repair. However, there are a number of key remaining questions. First, what additional tissues rely on Schwann cell recruitment for successful repair? Do Schwann cell growth factors play important roles in vasculature or organ repair, for example? Evidence that this may be the case comes from a consideration of the wide range of bioactive ligands expressed by dedifferentiated Schwann cells (Toma et al. 2020). Second, are all Schwann cells equal in their ability to promote repair and/or are their reparative abilities modulated by the tissue environment? Are Schwann cells in the regenerating digit tip the same or distinct from those in the skin when scarring is occurring? Finally, because transplanted Schwann cells could enhance murine digit tip regeneration, then can they also promote repair of tissues in which healing does not occur well, such as in diabetic skin ulcers?

In addition to Schwann cell paracrine mechanisms, the work described here indicates that nerve-derived mesenchymal cells contribute cells for mesenchymal tissue repair. Although this finding is unexpected, cells expressing PDGFRα with mesenchymal stem cell–like properties have been well described in many tissues (Farahani and Xaymardan 2015; Eisner et al. 2020; Santini et al. 2020). Nonetheless, these findings raise as many questions as they answer. For example, how similar are the neural crest–derived endoneurial mesenchymal cells to other adult mesenchymal precursor cell populations? How important are nerve-derived mesenchymal cells in tissue repair? Do they serve as a back-up plan or are they a star player? We also lack a detailed understanding of the molecular phenotype of the differentiated progeny that arise from nerve-derived mesenchymal cells. This question is particularly relevant because endoneurial mesenchymal cells are neural crest–derived and they contribute cells to mesodermally derived tissues. For example, normally healing skin is home to many heterogeneous fibroblast populations (Driskell et al. 2013; Shook et al. 2018; Mascharak et al. 2021). Do the endoneurial cells contribute a distinct fibroblast subpopulation? On a related note, within the injured nerve itself, do the endoneurial cells generate cells for other nerve compartments, such as the perineurium or even epineurial fibroblasts and adipocytes?

One final issue involves the potential clinical relevance and therapeutic utility of Schwann cells and nerve-derived mesenchymal cells for regenerative medicine strategies. Does dysfunction of either cell type contribute to chronic, nonhealing wounds like those seen in diabetes? Conversely, can these nerve-derived cells and/or their secreted factors be leveraged to enhance the endogenous repair of skin, bones, and/or other mesenchymal tissues? For example, could we convert a fibrotic scarring environment into a more regenerative environment by introducing dedifferentiated Schwann cells or the relevant growth factors? Answering these questions has the potential to lead to new therapeutic strategies in areas of broad clinical need.