Prior to discussing nerve coaptation techniques and the molecular mechanisms behind each of these, it is important to consider the general events occurring after nerve damage. Peripheral nerves consist of bundles of axons sheathed in three layers of connective tissue: epineurium, perineurium, and endoneurium (Figure 1). Each axon is enveloped by the endoneurium and Schwann Cells (SCs), the main glial cells of the peripheral nervous system. Groups of axons are organized into a fascicle by perineurium, and these fascicles are finally sheathed in epineurium to form the peripheral nerve. According to the Sunderland classification, peripheral nerve injuries are classified into five degrees depending on the mechanism and intensity of injury. These include segmental demyelination (first degree); axonal damage with intact endoneurium (second degree); axonal damage with endoneurium damage (third degree); axonal, endoneurial and perineurial damage (fourth degree); and loss of continuity of entire nerve trunk (fifth degree) [1,2]. For the purposes of this paper we will be discussing fifth degree nerve injury and regeneration. Damage to the axonal membrane results in two distinct temporal phases, a rapid phase consisting of rapid influx of sodium and calcium, and a slow phase characterized by activation and the retrograde transport of signaling molecules. The rapid influx of sodium and calcium activates calpain, leading to the sealing of the injured nerve membranes. It also generates a bust of action potentials at the proximal stump that is propagated to the soma via L-type voltage-gated calcium channels and the release of calcium from the endoplasmic reticulum. This results in the activation of the second messenger cAMP, which activates several important pro-regenerative transcription factors including dual leucine zipper-bearing kinase (DKL; also known as MAP3K12) and CREB1 (cAMP-responsive element-binding protein 1) [7,8,9,10]. The slower phase of injury is mediated via protein kinases, including calcium/calmodulin-dependent kinase 2(CMAK2); mitogen-activated protein kinases (MAPK) such as Erk1 and Erk2; phosphatidylinositol 3-kinase (PI3K); protein kinase A (PKA); protein kinase C (PKC); and c-jun N-terminal kinase (JNK). Retrograde trafficking systems transport these to the cell soma which then activates transcription factors such as CREB (cAMP responsive element binding protein), c-Jun, and ATF-3 (activating transcription factor 3). Combination of these processes enable activation of regeneration associated genes (RAGs) converting the nerve from a neuro-transmitter state to a regenerative state. For instance, downstream of the injury-induced PI3K-GSK3 (glycogen synthase kinase-3 signaling pathway), upregulates and phosphorylates SMAD family member 1 (SMAD1), causing its accumulation in the nucleus where it interacts with histone acetyltransferase p300 (p300HAT) which causes the over-expression of genes encoding cytoskeleton proteins such as tubulin and actin. Additionally, transcription factors (Sox11, c-JUN); cell adhesion and guidance receptors; growth-associated protein-43 (GAP43); and neurotropic factors (NTFs) such as NGF (Nerve Growth Factor) and BDNF (Brain-Derived Neurotropic Factor) and their receptors Trk, are also upregulated. There is a simultaneous downregulation of neurofilaments, neuropeptides and neurotransmitters gene expression, thus inhibiting the neurotransmitter function of the neuron. It also results in the ‘cell body response’ that is seen after PNI and includes swelling of the neuronal cell; chromatolysis (scattering of Nissl bodies or rough endoplasmic reticulum); migration of the nucleus to the peripheral of the neuron; and increased protein synthesis for regeneration [7,8,9,10,11,12,13,14,15].

Distally, the axons and myelin in the distal stump degenerate and are cleared (via a cascade of molecular and cellular events known as Wallerian degeneration (WD) (Figure 2) [16,17]. WD is a crucial process as the debris generates inhibitory signals such as myelin-associated glycoprotein that prevent axonal nascent neural growth from the proximal stump. SCs play a key role in the distal segment during WD. Axonal degeneration causes the loss of axonal contact with their SCs causing these adaptable cells (i.e., SCs) to re-enter the cell cycle and de-differentiate from a supportive to a repair phenotype. Gene expression of structural proteins such as myelin basic protein and myelin-associated glycoprotein are reduced, while reparative proteins such as NGF, basic fibroblast growth factor, Neurotrophin-3 (NT-3), neural cell adhesion molecule (NCAM), and glial maturation factor-β gene expression is upregulated. The de-differentiated SCs phagocytize degenerated myelin and repurpose it for future remyelination. They also express chemokine C-C motif ligand 2 (CCL2), leukemia inhibitory factor (LIF), IL-1α, IL-1β, IL-6, and pancreatitis-associated protein III (PAP-III) through the leaky nerve-blood barrier to recruit immune cells for further debris clearance [13,14,15,16,17]. Among immune cells, the first responders to the injured area are neutrophils, followed by macrophages that arrive 2–3 days after injury and peak at one week [16]. Macrophages clean up myelin via phospholipase A2 (PLA2) initiated phagocytosis, whose products, lysophosphatidylcholine (LPC) and C-reactive protein, activate the classic complement pathway [16]. In addition to cleaning up the debris, macrophages contribute to neural regeneration by producing proteases and growth-promoting factors, stimulating ECM remodeling, and regulating SCs. In addition to cleaning up the debris, macrophages contribute to neural regeneration by producing proteases and growth-promoting factors, stimulating ECM remodeling, and regulating Schwann cells [16].

Early peripheral regeneration studies by Cruikshank in the late 18th century were instrumental in the ultimate surgical repair of nerves. Von Langebeck performed one of the earliest successful surgical nerve repairs of the median nerve in 1876, with return of function in a year [2,18,19]. Woolsey was first to describe the end-to-end suture technique in 1907 which led to the end-to-end (ETE) surgical repair being more acceptable in clinical practice [18,19]. World War I accelerated the developments in the field of nerve repair and yielded procedures such as nerve grafting, described by Tinel and Elsberg [16]. As nerve grafting became more popular, Tinel, Elsberg, and Babcock detailed surgical decision making and when end-to-end repair is preferred to grafts [16]. In 1964, Smith popularized the use of the microscope for peripheral nerve surgery, which enabled more advanced nerve repairs and led to the modern surgical nerve repair techniques [18,19].

Primary end-to-end neurorrhaphy (ETE), is the current standard for nerve repair if the repair can be performed in a tension-free manner [18]. If a tension-free primary repair is not possible a nerve graft (autograft vs. allograft) is utilized to bridge the gap and coapted to the proximal and distal nerve stumps using the ETE technique. Several end-to-end repair techniques have been attempted, including epineural, fascicular, and group-fascicular repair (Figure 3) [2]. Epineurial repair involves placing microsutures through the epineurium without traumatizing the fascicles. It has the advantage of reducing suture material between the two ends and therefore does not hinder axonal regeneration. It is also technically easier to perform and does not damage axons. Fascicular repair attempts to minimize the misdirection of fibers by allowing for direct coaptation of fascicles, but is technically extremely challenging, requires a greater number of microsutures through the perineurium, and has a higher chance of axonal damage. The additional sutures also tend to cause more foreign body reaction and scarring, hindering axonal regeneration. Group-fascicular repair entails repair of groups of fascicles rather than single fascicles and reduces fascicular injury and scarring, however it still creates additional tissue damage and foreign body reaction with additional sutures when compared to epineural repair [2,3]. Thus, the epineural repair (Figure 3A) has emerged the current gold standard for ETE nerve coaptation since it has the least interfascicular foreign body reaction, least scaring and best functional outcomes of these three techniques [2,18,19,20].

Timing and distance of the nerve injury from the target muscle are extremely important considerations while repairing peripheral nerve injuries. In general, the optimal timing for performing nerve repair is generally accepted to be within 72 h of injury, with acceptable outcomes up to 7 days after injury [2,21]. There are three time-dependent events that occur after nerve injury: (1) Wallerian degeneration; (2) the rate of nerve regeneration; and (3) motor end plate (MEP) degeneration that occurs after prolonged denervation. Axonal regeneration only commences once Wallerian degeneration is completed which takes 3–4 weeks to be completed. Once WD is complete, axonal regeneration begins at the rate of 1 mm/day or 1 inch/month [13]. Additionally, MEPs irreversibly degenerate at 12–18 months after injury if not reinnervated by then, due to the loss of nerve stimulation. Hence with proximal nerve injuries that are over 18 inches proximal to the target muscle (for instance neck injuries or brachial plexus injuries), even if the nerve coaptation is performed immediately after injury, it is destined to fail since the regenerating axons will only reach the MEPs after 18 months, by which time the MEPs have already irreversibly degenerated. Similarly, if the repair is performed after 12–18 months, there would not be any effective motor functional recovery, making it imperative to repair nerve injuries as soon as possible, especially in proximal injuries [2,13,20,21]. Hence for proximal injuries or delayed presentations, nerve transfers are utilized wherein a donor nerve (from a less useful synergistic muscle) closer to the denervated (but more useful) muscle is coapted primarily (i.e., ETE) to rescue denervated MEPs. This does require the sacrifice of a healthy donor motor nerve, which denervates a healthy, albeit less useful muscle. Sensory nerves are far more forgiving with regards to return of function after delayed or proximal repairs. ETE nerve coaptations are the most utilized surgical technique for PNIs clinically, however, there are several important issues to consider while employing this technique. ETE works best with a tension-free anastomosis, i.e., the proximal and distal end can be brought together without any tension on the nerve. Additionally, it needs a proximal nerve stump to be available, which as discussed above, is not possible for nerve avulsion injuries where the nerves are avulsed from their nerve roots in the spinal cord.

In end-to-end repairs, the injured axon in the proximal nerve segment elongates giving rise to axonal sprouts with a highly mobile tip known as the growth cone. The growth cone can assume different shapes to survey their environment via membrane protrusions called filopodia and lamellipodia [20]. Growth cones guide the axons into the distal nerve segment along the gradient of neurotropic cues in the microenvironment, including extracellular matrix, guidance receptors, cell adhesion molecules and NTFs [21,22]. Axonal growth thus occurs in three stages: protrusion, driven by filamentous actin (F-actin); engorgement, led by microtubule transport of organelles into the region; and consolidation, which occurs when the proximal growth cone is stabilized [23,24]. This process is tightly governed by molecular mechanisms that we explore in this section and are summarized in Figure 4. As previously mentioned, neuronal injury triggers a series of signals that ultimately alter gene expression within the injured neurons [11]. One mechanistic component involves the downregulation of myelin-differentiation genes: myelin transcription factor Egr2 (Krox20), cholesterol synthesis enzymes, P0 structural protein, myelin basic protein (MBP), and membrane associated proteins, such as myelin associated glycoprotein (MAG) and periaxin [25,26]. These steps are important to remove inhibitory signals, allowing the axonal sprouts to reach their distal sites. Subsequently, genes that are involved in the recruitment of pre-myelinating Schwann cells are up-regulated, notably L1, neural cell adhesion molecules (CAMs), p75 neurotrophin receptor (p75NTR), and glial fibrillary acidic protein (GFAP) [25,26,27]. Upregulation of RAGs (also known as growth-associated genes) in SCs is transient but is another essential step that includes secretion of NTFs, such as BDNF, glial derived neurotrophic factor (GDNF), nerve growth factor (NGF), artemin (Artn), neurotrophin-3 (NT3), VEGF, and ciliary neurotrophic factors (CNTF) (Figure 4) [11,26,27,28,29,30]. Of note, CNTF promotes phosphorylation of transcription factor STAT3 which is retrogradely transported to the neuron soma and is essential for axonal regeneration and neuronal protection against apoptosis. In addition to neurotrophic factors, cytoskeletal proteins undergo qualitative and quantitative changes characterized by an increase in actin, peripherin and tubulin that are essential for growth cone adhesion, and a decrease in neurofilaments. Basal lamina proteins, such as laminin allow SCs to interact with the growth cone adaptor molecules to promote axonal growth into the endoneurial tubes of the distal nerve stump following end-to-end repair [7,12,31,32]. Clinically, the IKVAV motif of laminin has been used in stem cell therapy for peripheral nerve injury due to its ability to guide SC migration [33,34,35]. Changes in SC gene expression are controlled by factors like c-Jun, which induces a cascade involving a transient upregulation of cyclin-dependent kinase Cdc2, in turn phosphorylating vimentin and allowing it to interact with basal lamina proteins [12,36,37,38,39]. Guidance molecules such a Netrin-1, Ephrin B2 and Slit3 are also upregulated at the injury site within the fibrous bridge that forms across the coaptation site after ETE coaptation. Macrophages secrete high levels of Slit3, while fibroblasts within the bridge and the advancing axons secrete Robo1 receptors their surface. The Slit3-Robo1 signaling axis acts as a restrictive signal to keep the axons within the nerve bridge. Netrin-1 expression on SCs has a dual action on endothelial cells (ECs) and the growth cone. It interacts with CD146 receptors on ECs, boosting their proliferation and migration, while it binds with DCC receptor (Deleted in Colorectal Cancer) on advancing axons, guiding the growth cone across the injury site [35].

Thus, extracellular matrix proteins, NTFs, and RAGs expressed by SCs are crucial for successful axonal growth, and if any component is missing, regeneration will not occur [22]. Notably, these changes in cellular expression are transient, and declines further if neurons are not in contact with their target, the distal segment [16]. This has significant clinical implications. Firstly, it explains why surgical nerve coaptations need to be performed within 72 h if possible. Secondly, it is essential to freshen the nerve ends prior to coaptation so as to prime the neurons to express the optimal RAGs. Freshening the nerve endings also helps tip the balance between the M1 and M2 macrophage polarization towards nerve regeneration. A shift towards M1 state and a lack of M2 macrophages leads to chronic inflammation and impaired nerve regeneration. M2 macrophages are extremely important to nerve regeneration due to their ability to response to injury-induced hypoxia. Hypoxia triggers hypoxia-inducible factor 1 alpha (HIF-1α) production in macrophages which is followed by an increased expression of VEGF-A (Vascular Endothelial growth factor A). This in turn increases EC migration and angiogenesis across the coaptation site, that guide SCs across to form bands of Büngner across the injury site. These bands are crucial for axonal guidance as axons march across the gap from the proximal to the distal nerve segment [16].

4. Muscle Reinnervation