Together, these data summarized in 1–6 above strongly support the basic scientific concept and clinical promise of PEG-fusion for repairing singly cut PNIs—and there is good reason to believe data in 1–6 above will hold for clinical use of PEG-fused PNIs to repair segmental-loss PNIs.
PEG-fused PNAs to repair 0.5–1.0 cm segmental-loss PNIs in rat sciatic nerves exhibit atypical synaptic and other CNS and PNS plasticities to restore voluntary behaviors
After segmental-loss PNIs of 5–10 mm lengths in host Sprague Dawley rat sciatic nerves, PEG-fused PNAs from other donor wild type Sprague Dawley rats restored lost behavioral functions within 14–42-day PO (Fig.
3D) when neither host nor donors were tissue-matched or immune-suppressed. This behavioral restoration was almost-certainly by extensive PNS and CNS synaptic plasticities and collateral outgrowths, some of which are typically observed in adult mammals and others more typically restricted to embryonic stages of neuromuscular growth and innervation PO [
4,
14,
15,
30,
32,
33,
66]. For example, PEG-fusion immediately preserves spinal motoneurons, changes their peripheral connectivity, and alters dendritic organization (Fig.
4). This spinal reorganization may contribute to the remarkable behavioral recovery that is not present at the time of axonal repair, but develops in the following weeks. CNS and PNS post-natal behavioral recoveries in response to PNS mis-wirings occur by training or exercise regimens and environmental enrichments [
117]. Extreme levels of possible CNS rearrangements have been studied in rotations of sensory connections from the back and belly skin in frogs to produce changes in voluntary behaviors [
118]. All these data suggest that spinal and supra-spinal CNS plasticities can play a role in pattern relearning for motoneurons that survive PNIs—and provide evidence of alternate pathways for restoration of function independent of reinnervation specificity.
In addition to these CNS plasticities, alterations in peripheral synapses (PNS plasticities) were also commonly observed after PEG-fusion repair. NMJs after sciatic severance or ablation followed by PEG fusion exhibited changes consistent with an environment of partial denervation and increased synaptic activity [
4]. While some NMJs remained normally innervated (Fig.
2H–J), Terminal Schwann cell activation and process extension were seen in denervated NMJs at early timepoints (7–21 days). Early hyperinnervation followed by synapse elimination was also commonly observed after PEG-fusion, consistent with an environment of increased synaptic activity commonly seen after partial denervation. In contrast, no muscle fibers in NC PNAs were innervated from 7 to 21 days post-operatively, and
very few NMJs were innervated at 42 days post-operatively compared to almost 100% innervation in PEG-fused PNAs (Table
2) [
4].
PEG-fusion repair of some axons in a donor PNA and host distal nerve does not prevent natural regeneration of non-fused axons by outgrowths from surviving host stumps proximal to the PEG-fused PNA for at least 42-day post-injury. However, current data show that such outgrowths add little or nothing to the recovery obtained by surviving PEG-fused axons in PNAs at 42-day PO (Fig.
3D), and perhaps for many weeks thereafter in a few animals studied for longer PO times [
4]. It is possible that such a second wave of regenerating outgrowths might be obscured by continued improvement from PEG-fused axons due to CNS and/or PNS plasticities and collateralization of fused axons.
Assessing the contribution of collateralization to functional recovery in longer term studies is complex. In peripheral nerve regeneration without axonal fusion, recovery of function is due to two major axonal processes: axonal outgrowth and collateralization of regenerated/ing axons. Axonal fusion adds at least two additional processes: fused axons and collateralization of fused axons. Unfortunately, there is no easy method to definitively disentangle how each of these four processes individually contribute to functional recovery, further complicated by the several branch points distal to a PNA and the possible contributions of central/peripheral plasticities.
In brief, PEG-fusion of PNAs must produce its dramatic functional/behavioral recovery by activating peripheral and CNS synaptic and other plasticities, quite possibly to a much greater extent than most neuroscientists currently believe to be possible [
3,
30,
66,
119]. As noted [
3,
4,
7,
14,
15], this systems level adaptation may include a profound reorganization that restores a normal pattern of behavior (‘multiple realizability’ [
117]) and de-repress genes normally not active beyond embryonic or early postnatal stages. Trophic substances in the periphery, spinal cord and higher brain centers may help direct such re-organizations. Finally, extensive mis-connections in PEG-fused PNAs may evoke extensive synaptic plasticities and CNS and PNS axonal or dendritic outgrowths and rewiring to produce extensive behavioral recoveries.
PEG-fused PNAs that repair 0.5–1.0 cm segmental-loss PNIs in rat sciatic nerves exhibit atypical immunological properties
Two recent studies [
14,
15] have demonstrated that successfully PEG-fused PNAs that are not tissue-matched or treated with systemic immunosuppressive drugs exhibit an
atypical immunosuppressive microenvironment (Fig.
5A, B). Morphological, functional, immunohistochemical, and transcriptional analyses showed that many innate and adaptive immune responses that typically reject allogenic tissues were significantly attenuated in PEG-fused PNAs and do not produce functional rejection.
TEM data showed that PEG-fused PNAs maintained many large diameter axons (> 3 µm) that were well-myelinated by Schwann cells from 7- to 42-day PO [
4,
14,
15]. The persistence of myelinating Schwann cells within PEG-fused PNAs suggested that many donor Schwann cells were not rejected at any post-operative (PO) time from 0 to 42 days. In contrast to NC PNAs that were not treated with PEG and that were rejected within 14–21-day PO, PEG-fused PNAs did not exhibit collapsed hollow Schwann cell basal laminae or degradation of epineural sheaths and blood vessel basal laminae at 21-day PO. lHC analyses showed that T cell and macrophage infiltration, phagocytic activity, and expression of major histocompatibility complex (MHC) I and II glycoproteins necessary for antigen presentation were significantly reduced in the intra-fascicular mid-graft regions of PEG-fused PNAs by 21-day PO [
15]. In addition, PEG-fused PNAs had consistently reduced apoptotic activity as assessed via cleaved Caspase 3 immunostaining from 7- to 21-day PO (Fig.
5A).
Host T cells require chemotactic signals to infiltrate allograft tissues, and a combination of antigen interactions, cytokines, and co-stimulation to fully activate in response to donor cells [
35]. Compared to NC PNAs, PEG-fused PNAs had consistently low T cell infiltration (Fig.
5), significantly lower at 14-day PO [
15]. The reduction in T cell infiltration coincided with significantly reduced gene expression of key cytokines and chemokines involved in T cell trafficking and Th1 effector cell polarization, such as C–X–C Motif Chemokine Ligand 11 (CXCL11) and Interferon Gamma (IFN-γ) in PEG-fused PNAs [
15,
120,
121]. An exploratory, bulk, non-specific, RNA sequencing study of the coding transcriptome of PEG-fused PNAs at 14-day PO showed molecular processes consistent with immunotolerance [
14]. That is, compared to NC PNAs and Unoperated Control nerves, PEG-fused PNAs significant downregulated many T cell-associated gene transcripts (Fig.
5), including: (1) costimulatory receptors such as CD28; (2) Interleukin 2 Receptor Alpha (IL2RA); (3) cytokines involved in inflammatory Th1 cell polarization such as Interleukin 2 (IL-2) and Interleukin 12B (IL-12B); and, (4) transcriptional factors necessary for T cell activation such as GATA3 and T-box 21 (TBX21) [
14,
122‐
125]. These results suggest either direct or indirect inhibition of T cell migration as well as activation in PEG-fused PNAs, which might be produced through several different mechanisms.
Reduced T cell activity might be due to an “immunocamouflage” effect produced in donor axon-Schwann cell units that have received host proteins and/or MHC molecules as a result of PEG-fusion. Nearby host T cells as well as host antigen presenting cells might not respond to axon-Schwann cell units that present host antigens. Direct inhibition of inflammatory T cell activation and effector functions by suppressor cells such as Tregs that secrete the anti-inflammatory cytokine IL-10 might evoke a local immune-acceptance in PEG-fused PNAs [
108]. Lack of CD28 co-stimulation in host T cells upon antigen recognition might render them anergic in PEG-fused PNAs, thereby preventing them from responding properly [
126].
However, as summarized in Smith et al., 2020B [
14] and Fig.
5B, transcripts commonly involved in Th1 cell suppression such as IL-10, Cytotoxic T-lymphocyte Associated Protein 4 (CTLA4), and CD274 (also known as Programmed Death Ligand 1 (PD-L1)) were also significantly downregulated [
14]. CTLA4 and PD-L1 are immune checkpoint inhibitors that bind to CD28 and CD80, respectively, on T cells and are critical therapeutic targets for immunosuppression via costimulatory blockade [
127,
128]. The significant downregulation of these transcripts suggested that they were not heavily involved in immunosuppression associated with PEG-fusion. Nonetheless, CTLA4 and PD-L1-mediated immunosuppression of T cell activation pathways in PEG-fused PNAs is an important potential mechanism that warrants further investigation at the protein level.
Macrophages are numerous in injured peripheral nerves and PNAs [
69]. In addition to their role in debris phagocytosis and promotion of regenerative processes in peripheral nerves, macrophages also heavily influence the inflammatory environment and often contribute to allograft rejection [
35]. Upon tissue injury, many macrophages initially adopt a pro-inflammatory “M1” activation state in which they produce inflammatory cytokines, such as IL-12, and high levels of nitric oxide (NO) that can damage nearby cells [
129,
130]. M1 polarization is stimulated by IFN-γ from Th1 cells, while IL-12 partially drives Th1 cell polarization [
122,
130]. Macrophages thereby engage in positive feedback loop crosstalk with T cells to promote inflammation, antigen recognition, and donor cell death. In immunotolerated tissues, the reverse situation is also possible in which anti-inflammatory “M2” macrophages and Tregs can induce each other’s activation states to suppress Th1- and M1-mediated inflammatory responses and promote wound healing [
108,
131,
132].
PEG-fused PNAs in IHC studies (Fig.
5A) had significantly reduced macrophage infiltration and phagocytic activity via CD68 and IBA1 immunostaining by 14–21-day PO compared to NC PNAs [
15]. At 14-day PO, this result coincided with significantly reduced expression of common M1 macrophage markers such as inducible nitric oxide synthase (iNOS), reduced expression of the inflammatory cytokine Interleukin 1β (IL-1β) often produced by macrophages, and increased expression of common M2 macrophage markers, such as Arginase 1 (ARG1) and Mannose Receptor C-type 1 (MRC1, also known as CD206) [
14,
130,
132]. These results suggest that PEG-fused PNAs might have both a reduction in macrophage infiltration as well as a shift in macrophage polarization that may contribute to an immunosuppressive environment. It is possible that lack of axonal damage in PEG-fused PNAs produces alternative Schwann cell signaling mechanisms and, therefore, alternative macrophage polarization.
The cleaved form of Caspase 3 is a common executioner of Granzyme- and Fas-mediated apoptosis by cytotoxic CD8 T cells or natural killer (NK) cells in rejected non-neuronal allografts and PNAs [
35,
133‐
135]. Superoxide and NO cytotoxicity from ischemia reperfusion injury in organ allografts have also been shown to induce apoptosis via Caspase 3 [
136]. After single-transection injury in sciatic nerves, Schwann cells, macrophages, neutrophils, and other cells regularly undergo turnover via apoptosis throughout the degeneration and regeneration process [
69,
71]. Compared to NC PNAs, cleaved Caspase 3 immunostaining was consistently reduced to near-unoperated nerve levels in PEG-fused PNAs by 21-day PO, while Granzyme B and FasL expression were reduced on average at 14-day PO [
15]. These data (Fig.
5) [
15] suggested that both WD-associated apoptosis and rejection-associated apoptosis were inhibited in PEG-fused PNAs and might contribute to their immune-acceptance and functional non-rejection.
MHCI/II protein expression was reduced in PEG-fused PNAs and a set of transcripts encoding essential antigen presentation machinery components were downregulated in PEG-fused PNAs [
14,
15]. These downregulated transcripts included Transporter 2 (TAP2) protein, needed to transport antigen peptides from the cytoplasm to the endoplasmic reticulum [
137], and transcription factors that drive MHCI expression [Nod-like Receptor C5 (NLRC5)] and MHCII expression [Class II Major Histocompatibility Complex Transactivator (CIITA)] [
138,
139]. These results (Fig.
5) [
14,
15] suggested that mechanisms for antigen presentation and immunogenicity were also suppressed in PEG-fused PNAs. Lisak et al. (2016) [
72] demonstrated that MHCII expression in SCs depends on their myelination state [
72]. Myelinating SCs do not express MHCII and do not respond to IFN-γ stimulation, while demyelinated SCs do upregulate MHCII [
72]. Therefore, the maintenance of many viable myelinated axons in PEG-fused PNAs may inhibit SC expression of MHCII and render those SCs non-immunogenic. Similar mechanisms might be involved in reduced MHCI expression of SCs in PEG-fused PNAs. It is also possible that the soluble 3.35 kD PEG used to PEG-fuse PNAs [
14,
15] in these studies could inhibit MHC expression in PNAs, as shown in other studies examining kidney allograft storage in organ preservation solutions containing soluble PEG before transplantation [
140].
Although PEG-fused PNAs had reduced immune responses described above relative to NC PNAs (Fig.
5A), PEG-fused PNA immune responses were much greater than these immune responses in Unoperated Control sciatic nerves [
15]. Cellular infiltration, MHC protein expression, and cytokine/chemokine expression were higher in PEG-fused PNAs, suggesting that the signals necessary for orchestrating innate and adaptive immune responses were not completely eliminated, but rather suppressed at the transcriptional or post-translational level. It is possible that physical barriers in the blood vasculature, basal laminae, or other extracellular matrix components in PEG-fused PNAs impede infiltration by those immune cells, and thereby limit donor or host immune cell activation or induce alternative activation of immunosuppressive T cell and/or macrophage phenotypes [
141‐
144]. Future studies will require single-cell sequencing, flow cytometry, and additional in vivo analyses to provide greater clarity as to the cell-specific nature of differential gene expression, cytokine production, and specific regulatory mechanisms driving immunosuppression in PEG-fused PNAs.