Cell and Tissue Research (2022) 390:355–366 https://doi.org/10.1007/s00441-022-03682-1 REGULAR ARTICLE A novel decellularized nerve graft for repairing peripheral nerve long gap injury in the rat Estefanía Contreras1 · Sara Bolívar1 · Núria Nieto‑Nicolau2,3 · Oscar Fariñas2,3 · Patrícia López‑Chicón2,3 · Xavier Navarro1 · Esther Udina1 Received: 2 June 2022 / Accepted: 26 August 2022 / Published online: 17 September 2022 © The Author(s) 2022, corrected publication 2023 Abstract Decellularized nerve allografts are an alternative to autograft for repairing severe nerve injuries, since they have higher availability and do not induce rejection. In this study, we have assessed the regenerative potential of a novel decellularization protocol for human and rat nerves for repairing nerve resections, compared to the gold standard autograft. A 15-mm gap in the sciatic nerve was repaired with decellularized rat allograft (DC-RA), decellularized human xenograft (DC-HX), or fresh autograft (AG). Electrophysiology tests were performed monthly to evaluate muscle reinnervation, whereas histologi- cal and immunohistochemical analyses of the grafts were evaluated at 4 months. A short-term study was also performed to compare the differences between the two decellularized grafts (DC-RA and DC-HX) in early phases of regeneration. The decellularization process eliminated cellularity while preserving the ECM and endoneurial tubules of both rat and human nerves. Higher amount of reinnervation was observed in the AG group compared to the DC-RA group, while only half of the animals of the DC-HX showed distal muscle reinnervation. The density of myelinated axons was significantly higher in AG compared to both DC grafts, being this density significantly higher in DC-RA than in DC-HX. At short term, fibroblasts repopulated the DC-RA graft, supporting regenerated axons, whereas an important fibrotic reaction was observed around DC-HX grafts. In conclusion, the decellularized allograft sustained regeneration through a long gap in the rat although at a slower rate compared to the ideal autograft, whereas regeneration was limited or even failed when using a decellularized xenograft. Keywords Nerve regeneration · Decellularization · Nerve allograft · Nerve injury Introduction the EU per year. Although peripheral neurons have the abil- ity to regenerate their axons and eventually reinnervate the Peripheral nerve injury results in partial or total loss of the previously denervated target organs, clinical and experimen- sensory and motor functions dependent on the injured nerve(s), tal evidence shows that regeneration is often unsatisfactory, with important consequences for the quality of life of affected especially following severe injuries (Pfister et al. 2011; Lovati subjects (Navarro et al. 2007). It is estimated an incidence of et al. 2018). nerve injuries of 13.9/100,000 inhabitants per year (Asplund Surgical repair is needed after nerve transections, to re- et al. 2009) resulting in more than 300,000 such injuries in unite the two nerve stumps, with the aim to facilitate proxi- mal axons to regenerate through the distal degenerating nerve and to finally reinnervate their target organs (Pfister * Esther Udina esther.udina@uab.cat et al. 2011). However, direct suture is not always possible, and long nerve gaps must be bridged. The clinical gold 1 Department of Cell Biology, Physiology and Immunology, standard repair technic in clinics is the interposition of an Institute of Neuroscience, Universitat Autònoma de autologous nerve graft between the proximal and the dis- Barcelona, and CIBERNED, ISCIII, 08913 Bellaterra, Spain tal stump (Grinsell and Keating 2014; Gaudin et al. 2016; 2 Barcelona Tissue Bank, Banc de Sang I Teixits (BST), Kornfeld et al. 2021). Although it sustains nerve regen- Barcelona, Spain eration across long defects, the use of an autografts has 3 Biomedical Research Institute (IIB-Sant Pau; SGR1113), some disadvantages, such as limited availability of donor Barcelona, Spain 13 Vol.:(0123456789) 356 Cell and Tissue Research (2022) 390:355–366 nerves, increased operating time, and morbidity at the site Regarding animal experiments, all the procedures were of extraction (e.g., pain, scars, neuroma, and sensory loss). approved by the Ethics Committee of the Universitat Autònoma Although devoid of the drawbacks of autografts, allografts de Barcelona and Generalitat de Catalunya (reference #10,306) induce an immune rejection response by the recipient body and followed the European Community Council Directive and therefore require systemic immunosuppressive therapy. 2010/63/EU of the European Parliament on the protection of The most immunogenic elements in the allografts are Schwann animals used for scientific purposes. In addition, we followed cells and myelin sheaths, since their membrane present major the ARRIVE guidelines and committed ourselves to the 3Rs of histocompatibility complex antigens (Evans et al. 1995). laboratory animal research. Immune compatibility between donor and host is thus impor- tant to guarantee axonal regeneration, in both nerve grafts and Preparation of acellular nerve grafts artificial guides pre-filled with transplanted Schwann cells (Rodríguez et al. 2000). A promising alternative to autografts Human sural nerves from eleven deceased donors were is a decellularized nerve allograft (Hundepool et al. 2017; obtained from the Barcelona Tissue Bank-Banc de Sang i Lovati et al. 2018; Philips et al. 2018). The decellularization Teixits (BTB-BST, Barcelona, Spain; https://w​ ww.b​ ancsa​ ng.​ involves several processes that guarantee immunogenic free net/) within 24 h post-mortem. Donor screening included, scaffolds from native nerves, while preserving the extracellular but may not be limited to, the review of complete social and matrix (ECM) components and the biomechanical properties medical history, physical examination of the donor, complete of the grafts (Szynkaruk et al. 2013; Nieto-Nicolau et al. 2021). serological and microbiological testing during retrieval, his- Therefore, decellularization of human peripheral nerves is an topathological analysis, and any other information pertaining innovative strategy in tissue engineering and it can become an to risk factors for relevant communicable diseases. Sural alternative to autografts to repair long-gap peripheral nerve nerve fragments were retrieved from the donor and were injuries. For its clinical translation, it is mandatory to evalu- placed into a sterile container with Roswell Park Memorial ate the pro-regenerative potential of optimized protocols of Institute (RPMI) media without phenol (Gibco, Carlsbad, decellularization of human cadaver nerves. However, the pre- CA, USA) plus 1% antibiotics (vancomycin [Pfizer, MA, clinical evaluation of human grafts in experimental models has Spain], penicillin [Normon, MA, Spain], and streptomycin an important limitation, since these grafts become xenografts [Reig Jofre, BCN, Spain]) at 4 °C until its decellularization when transplanted to experimental animals, and the immune in a clean room environment. rejection caused by donor antigens (Udina et al. 2003) or other Rat sciatic nerves were harvested from Sprague–Dawley inter-species differences (Wood et al. 2014) can interfere with donor rats, under deep anesthesia and with sterile conditions. their regenerative potential. Nevertheless, it is worth to further Samples were placed in sterile phosphate buffered solution explore the potential of decellularized xenografts as a repair (PBS) with antibiotic/antifungal agents (Sigma-Aldrich) and alternative to grafts from the same species origin. Therefore, processed within 24 h. in this study, we aimed to perform a comparative assessment The decellularization protocol for human nerve samples of the effective axonal regeneration, along a critical nerve gap was as described previously (Nieto-Nicolau et al. 2022). To of 15 mm length induced in the sciatic nerve of adult rats, decellularize rat nerves, that protocol was adapted, repro- repaired with a decellularized allograft, a decellularized human ducing sequential incubations and using zwitterionic and xenograft, or the gold standard autograft. non-ionic detergents for 3 days. Zwitterionic sulfobetaines 10 and 16 were purchased from Sigma-Aldrich (Munich, Germany). Non-ionic Triton X-200 was changed by Triton X-100 (Sigma-Aldrich). This protocol was optimized adding Materials and methods an incubation with hypertonic 1 M NaCl (Sigma-Aldrich) during 4 h and 0.1 mg/ml Pulmozyme DNASE (Roche, Bar- Ethics approval celona, Spain) for 3 h. After DNASE treatment, the nerves were washed once in 0.5 M Tris–EDTA buffer and several This study followed the ethical precepts of the Declaration times in ultrapure water. The process was carried out at room of Helsinki and was approved by local ethics committee. temperature for 4 days. Human tissue was processed according to guidance for clini- cal use (EEC regulations 2004/23/CE and 2006/17/CE) and Histological evaluation of decellularized nerve grafts to the legal requirements for the use of biological samples for research in Spain (Law 14/2007 and RD 1716/2011). Decellularized nerve grafts were fixed in paraformaldehyde Ethical Committee approval was issued by CEIm Hospital 4% for 24 h at 4 °C, then transferred to phosphate buffered Valle Hebron, Barcelona; PR (BST) 314/2019. In all cases, (PB) and incubated with saccharose 30% for 24–48 h. Sam- informed consent was obtained from the donors’ relatives. ples were cryo-embedded with Tissue T ­ ek® OCT compound 13 Cell and Tissue Research (2022) 390:355–366 357 (Sakura, Fleminweg, Netherlands), and cross sections of 15 (TA), gastrocnemius (GA), and plantar interosseus (PL) µm thickness were obtained with a cryotome. After block- muscles. The reference electrode was placed at the fourth ing with a solution of normal goat serum or normal donkey toe, and a ground electrode was placed at the knee. The serum (10%) containing 0.3% Triton, sections were incu- amplitude and latency of the CMAP were measured. The bated overnight at 4 °C with primary antibodies against contralateral intact limbs were used as control. The rat body S-100 protein (S100; Schwann cells; 1:50; 22,520-DiaSorin), temperature was maintained throughout the test with a ther- neurofilament (NF200; myelinated axons; 1:400; AB5539- mostatic warming flat coil. Millipore), non-collagenous connective tissue glycoprotein (laminin; 1:500; AHP420-Biorad), and collagen (Collagen Histological evaluation IV; 1:400; 134,001-Setareh Biotec). Following washes, sec- tions were incubated with secondary antibodies bound to At the end of the 120-day follow-up, animals were eutha- Alexa Fluor 488 and Alexa Fluor 594. Sections were also nized by an intraperitoneal injection of pentobarbital, tran- stained with DAPI (DAPI; nuclei; 1:100; D9564-10MG- scardially perfused with 4% paraformaldehyde in PBS for Sigma) and myelin stain (Fluoromyelin; myelin; 1:300; 30 min, and the sciatic nerves were collected and divided F34651-Invitrogen). Immunolabeled sections were viewed in three parts. under epifluorescence microscopy (Olympus BX51). The middle segment of the nerve graft was processed for immunofluorescence labeling as indicated above. Serial sec- In vivo long‑term studies tions were incubated for staining axons (NF200), Schwann cells (S100), ionized calcium-binding adapter molecule 1 (Iba1; Female Sprague–Dawley rats (12 weeks of age and weighing macrophages; 1:500; 19–19,741-Rafer), and laminin. Samples 296–322 g) were used. They were housed under a 12-h-light–dark were washed and incubated with secondary antibodies Alexa cycle in a temperature-controlled animal care facility with ad libi- Fluor 488 goat anti-chicken (1:200; A11039-Invitrogen), Alexa tum access to water and food. Fluor 488 donkey anti-goat (1:200; A11055-Invitrogen), and Seventeen female Sprague–Dawley rats were randomized Alexa Fluor 594 goat anti-rabbit (1:200; A21207-Invitrogen) into three different groups: autograft (AG) (n = 6), decellular- diluted in PBS-Triton 0.3%. ized rat allograft (DC-RA) (n = 5), and decellularized human The proximal and distal segments of the grafted nerve, includ- xenograft (DC-HX) (n = 6), and followed for 120 days. Rats ing suture levels, were post-fixed in 3% glutaraldehyde −3% para- were anesthetized with ketamine (75 mg/kg) and medetomi- formaldehyde in cacodylate-buffer solution (0.1 m, pH 7.4) at dine (0.01 mg/kg) intraperitoneally. The right hindlimb was 4 °C. These segments were post-fixed with osmium tetroxide shaved and sterilized with povidone-iodine solution. A skin (2%, 2 h) and dehydrated through ethanol series prior to embed- incision was made, and the right sciatic nerve was exposed ding in epon resin. Semithin 0.5 μm thick sections were stained and resected. In the autograft group, a 15-mm-long nerve with toluidine blue and examined by light microscopy (Olym- segment of the sciatic nerve was excised and then, sutured pus BX40). Sets of images obtained at 1000 × were chosen by again in place using three 10–0 nylon epineural sutures. In the systematic sampling of squares representing at least 30% of the rat allograft and human xenograft, a 15-mm decellularized nerve cross-sectional area (Gómez et al. 1996). Measurements rat allograft or a 15-mm decellularized human xenograft was of cross-sectional area of the whole nerve and counts of the num- sutured to the host nerve stumps. The muscles were sutured ber of myelinated nerve fibers were carried out using ImageJ with 6–0 and the skin with 3–0 silk and staples. Postoperative software. care included amitriptyline (20 ml/l) in drinking water to prevent autotomy (Navarro et al. 1994) and buprenorphine 0.03 mg/kg subcutaneously to treat postoperative pain. In vivo short‑term studies Electrophysiological tests Twelve female Sprague–Dawley rats were randomized in two experimental groups of 6 animals each, according to the Reinnervation of target muscles was assessed at 30, 60, 90, repair procedure: DC-RA and DC-HX. In each group, half of and 120 days post-injury (dpi) by motor nerve conduction the animals were assessed at 7 days and half at 14 days post- tests. Animals were anesthetized as indicated above, and injury. Surgeries were performed as described above. How- the sciatic nerve was stimulated with transcutaneous needle ever, in this case, a segment of the sciatic nerve was excised electrodes placed at the sciatic notch delivering single pulses and substituted by a 12-mm DC-RA or 12-mm DC-HX, of increasing intensity (Synergy Medelec, Viasys Health- therefore using a shorter graft than for the long-term study. Care), and the compound muscle action potential (CMAP) The proximal and distal nerve stumps were sutured by 10–0 was recorded by placing electrodes on the tibialis anterior nylon epineural sutures. 13 358 Cell and Tissue Research (2022) 390:355–366 Histological evaluation Results At the end of the follow-up (7 or 14 dpi), animals were Effectiveness of the decellularization procedure euthanized and perfused as described. The sciatic nerve was harvested and divided in two equal parts, stored in cryo- Cross sections of human and rat decellularized nerve grafts protective solution of PBS-sucrose 30% at 4 °C for 24 h. were processed for immunohistochemistry. Immunolabeling The proximal nerve segment at 7 dpi and the proximal and against laminin and collagen IV indicated that the structure distal nerve segment at 14 dpi were embedded in Tissue of the extracellular matrix and endoneurial tubules was well Tek and 15-μm-thick longitudinal sections cut serially. For preserved and comparable. Neurofilament remnants were immunostaining, sections were incubated overnight with present though their concentration was highly reduced and primary antibodies for staining axons (NF200), Schwann fragmented, compatible with degradation of axons, whereas cells (S100), macrophages (Iba1), and fibroblasts (vimentin; the absence of DAPI and S100 staining indicated that cell 1:400; ab92547-Abcam). Samples were washed and incu- contents were completely removed (Fig. 1). bated with secondary antibodies as above. Finally, sections were washed and cover-slipped with Fluoromount contain- In vivo repair of a critical nerve gap ing DAPI (1:10,000, Sigma) for nuclear counterstain. with decellularized grafts In vitro study Electrophysiological test results Dissociate culture of DRG Nerve conduction tests were performed to assess reinner- vation of TA, GM, and PL muscles (Fig. 2). At 60 dpi, all Nine adult C57BL/6 J mice were euthanized, and DRGs animals from AG and DC-RA groups showed evidence of were dissected. Dissociation and culture of DRG cells were reinnervation in the TA and GM muscles, whereas in the PL performed as previously described (Alé et al. 2014). Briefly, muscle, CMAPs were recorded in all animals of AG group the ganglia were cleaned under a dissecting microscope and but only in 20% of DC-RA group. In the DC-HX rats, only enzymatically digested by incubation for 30 min at 37 °C in 50% of the animals showed evidence of reinnervation of 10% trypsin, 10% collagenase, and 10% DNAse. The enzy- the TA and GM muscles and none for the PL muscle. At matic activity was halted by adding 1 ml of DF10S, and 90 dpi, the amplitude of the CMAPs increased. All animals tissue suspension was centrifuged at 900 g for 3 min. Then, from AG and DC-RA groups had reinnervation of TA, GM, the pellet was resuspended in 1 mL of culture medium and and PL muscles, whereas half of the animals of the DC-HX mechanically dissociated. had positive values in TA and GM muscles and only one A total of 12,000 cells were plated on coverslips coated in the PL muscle. At the end of the follow-up (120 dpi), with poly-D-lysine plus human laminin (1 µg/µl, Sigma- the mean CMAP amplitude of TA (30.8 ± 1.0 mV), GM Aldrich) or mice laminin (1 µg/µl, Sigma-Aldrich). After 24 h (43.0 ± 2.6 mV), and PL (2.8 ± 0.4 mV) muscles in the AG in culture, cells were fixed with 4% paraformaldehyde for group were higher than in the DC-RA group (23.1 ± 3.7 mV, 20 min, and neurons were immunostained with β3-tubuline 31.4 ± 2.9 mV, and 1.5 ± 0.2 mV, respectively). Animals antibody (1:500, MMS-435P/801202, Biolegend), NF200 from DC-HX group showed significantly lower values com- antibody (1:800) and DAPI (1:1000, D9564-10MG-Sigma). pared to AG and DC-RA groups in the TA (9.7 ± 4.9 mV; The length of the longest neurite from β3-tubuline and p < 0.0001), GM (10.9 ± 5.7 mV; p < 0.01 vs DC-RA, NF200 positive neurons was measured at 100 × magnification and p < 0.0001 vS AG) and PL muscles (0.05 ± 0.05 mV; under epifluorescence microscopy. Three different cultures p < 0.001 vs DC-RA and p < 0.0001 vs AG). Moreover, only using both conditions in at least three different wells each half of the animals repaired with the human decellularized were analyzed. graft showed reinnervation signs in TA and GM muscles, and just 1 of the 6 animals in the PL muscle. Data analysis Histological nerve assessment Data are expressed as mean ± standard error of the mean (SEM). The results of functional tests and histology were At the end of follow-up, at 120 dpi, nerves were harvested analyzed by two-way ANOVA. The results of the longest and processed for histological analysis. There were numerous neurites from DRG cultures were compared using unpaired regenerated myelinated axons and Schwann cells within the t-test. Statistical analyses were made with GraphPad Prism AG and DC-RA grafts, but quite limited in DC-HX grafts. 8 software. A p < 0.05 was considered as significant. The endoneurial basal lamina tubules were preserved in all 13 Cell and Tissue Research (2022) 390:355–366 359 Fig. 1  Decellularization efficiency. Representative micrographs romyelin) (f″), in a control nerve (native nerve) (a, b, c, d, e, f), a showing immunolabeling of Schwann cells (s100) (a″), myelinated decellularized rat nerve allograft (a′, b′, c′, d′, e′, f′) and a decellular- axons (nf200) (b″), nuclei (dapi) (c″), extracellular matrix proteins ized human nerve xenograft (a″, b″, c″, d″, e″, f″). Images taken at (laminin and collagen IV) (d″ and e″, respectively), and myelin (fluo- 100 × magnification; scale bar 400 µm the grafts, and the amount of macrophage immunolabeling and ****p < 0.0001). All animals showed regeneration in was also similar (Fig. 3). the middle of the grafts. Distal to the nerve graft, all animals Quantitative analysis demonstrated that the density of from AG and DC-RA had myelinated axons, whereas only myelinated axons was statistically higher in AG (40267.78 ± 3/6 animals of the DC-HX showed positive results (Fig. 4). 2775.7 axons/mm2) than in the DC-RA (28132.47 ± 3084.11 axons/mm 2, *p<0.5 vs AG) and DC-HX (18244.98 ± Evaluation of short‑term regeneration 3070.41 axons/mm2, $p<0.5 vs DC-RA and ***p < 0.001 vs AG) groups in the middle of the graft. Distal to the graft, Considering the above results, a short-term study was per- the myelinated fiber density was also significantly higher formed to assess the early events of the regenerating nerves in AG group (27293.08 ± 195.8 axons/mm2) compared to in the two types of decellularized grafts. At 7 dpi, the axonal DC-RA (14337.13 ± 1324.54 axons/mm2, **p < 0.01) and regenerative front, labeled with NF200, reached the middle DC-HX (2289.29 ± 513.73 axons/mm2, #p<0.01 vs DC-RA of the graft in rats of both groups, although there was higher Fig. 2  Electrophysiological evaluation of nerve regeneration along (***p < 0.001 vs AG), tibialis anterior (**p < 0.01 vs AG; ***p < 0.001 the 120-day follow-up after sciatic nerve section and repair. Results vs DC-RA; ****p < 0.0001 vs AG; #p < 0.0001 vs DC-RA) and gas- are presented as mean ± SEM. Statistical analysis was performed trocnemius (*p < 0.5 vs AG; #p < 0.5 vs AG; **p < 0.01 vs DC-RA; using 2-way ANOVA. Plots show the amplitude of CMAPs of plantar ***p < 0.001 vs AG; ****p < 0.0001 vs AG) muscles 13 360 Cell and Tissue Research (2022) 390:355–366 Fig. 3  Immunohistochemical characterization of the grafts 120 days teins labeled against laminin (a″, b″, c″, d″), and macrophages after implantation. Representative micrographs of nerve grafts har- labeled against Iba1 (a‴, b‴, c‴, d‴) in cross sections of the graft in vested at 120 dpi showing immunofluorescence of Schwann cells a control nerve (no decellularized native nerve) (a‴), autograft (b‴), labeled against S100 (a, b, c, d), myelinated axons labeled against rat allograft (c‴), and human xenograft (d‴) groups. Images taken at neurofilament 200 (NF200) (a′, b′, c′, d′), extracellular matrix pro- 200 × magnification; scale bar 150 µm number of axons in DC-RA than in DC-HX grafts (Fig. 5a). animal that had already NF positive axons at the proximal There were relatively few Schwann cells accompanying the segment. As expected, in the other two animals, the graft regenerating axons. In the DC-RA group, marked prolifera- was disorganized, with infiltrating cells and marked presence tion of fibroblasts was observed within the nerve, whereas of macrophages. This is in contrast to what we observed at in the DC-HX group, the fibroblast invasion was mainly 7 days, when all animals of the group had some regenerative around the nerve. The Iba1 reactivity, corresponding to mac- axons in the proximal half of the graft. It is possible that the rophages, was higher in the DC-HX group (Fig. 5b). important fibrotic reaction developed at 14 dpi overcame the In samples taken at 14 dpi, a greater number of Schwann regenerative response observed at early time points. cells compared to 7 dpi was present around regenerating axons (Fig. 5c), whereas fibroblasts showed a more organ- Influence of ECM origin on neurite growth of DRG ized longitudinal organization. In the DC-HX group, fibro- neurons in culture blasts were present both inside and outside the graft, indi- cating an important fibrotic reaction, whereas macrophages To evaluate if inter-species differences in ECM components, were only present around the nerve (Fig. 5d). Besides, only specifically in laminin, could play a role in the poorer abil- one of the three animals presented regenerated axons along ity of decellularized xenografts to sustain axonal regenera- the nerve graft. In the DC-RA group, the regenerative front tion, we compared neurite growth of mice DRG neurons on reached the distal segment of the graft. In the DC-HX group, slides coated with laminin either from human or from mouse regenerating axons reached the distal segment in the only origin. When mice neurons were platted on mice laminin, 13 Cell and Tissue Research (2022) 390:355–366 361 Fig. 4  Histological evaluation of the regenerative potential of the magnification; scale bar 10 μm. (g) Plots showing density of myeli- grafts at 120 days. Representative transverse semithin sections of the nated axons in the sciatic nerve at the mid graft and distal to the graft mid graft (a, b, c) and distal to the graft (d, e, f) in autograft (a, d), in the three groups. *p<0.5 vs AG; $p<0.5 vs DC-RA; **p<0.01 vs decellularized rat allograft (b, e), and decellularized human xenograft AG; #p<0.01 vs DC-RA; ***p<0.001 vs AG; ****p<0.0001 vs AG (c, f) groups, stained with toluidine blue. Images were taken at × 1000 a higher number of neurons with neurites were observed Successful nerve regeneration requires an orchestrated series compared to human laminin. Mean of neurite length from of events that implies a switch of neurons to a pro-regenerative neurons with neurites was significantly higher in mouse state and important changes at the distal stump, that lead to laminin (175.7 ± 114.4 µm) compared to human laminin Wallerian degeneration and the creation of a permissive milieu (118.9 ± 66.9 µm, p < 0.0001). When analyzing neurite for axonal regeneration, where Schwann cells, immune cells, growth of myelinated (NF positive) and non-myelinated endothelial cells, and fibroblasts play an important role sup- (NF negative) neurons, similar results were found. There- porting axonal regrowth (Mcdonald et al. 2006; Allodi et al. fore, mean neurite length of NF positive neurons with neu- 2012; Kim et al. 2013; Dun and Parkinson 2015; Cattin and rites (185.7 ± 125.9 µm and 126.7 ± 66.5 µm, p < 0.0001) Lloyd 2016; Roballo and Bushman 2019). Despite the potential and NF negative neurons with neurites (157.9 ± 88.7 µm of peripheral nerves to regenerate, successful functional recov- and 106.28 ± 66.2 µm, p < 0.001) (Fig. 6) was significantly ery is usually limited after severe nerve lesions (Navarro et al. higher on mouse laminin compared to human laminin. 2007). Thus, nerve resections that lead to long gaps between nerve stumps present poor success of regeneration and com- monly limited recovery if surgically repaired (Pfister et al. 2011). This limitation is partially due to the lack of good repair Discussion alternatives to the gold standard autograft. Since it is mandatory to bridge the gap to guarantee some degree of regeneration, In this study, we have performed the first in vivo assay of research has focused on a variety of graft substitutes, including a novel, optimized protocol for nerve decellularization that decellularized graft (Nagao et al. 2011). A decellularized graft fulfills the standardized recommendations for peripheral has to provide an off-the-shelf alternative to synthetic conduits nerve decellularization (Nieto-Nicolau et al. 2021), and while partially maintaining some of the proregenerative proper- demonstrated that such decellularized allograft allows for ties of autologous nerve grafts (Kasper et al. 2020). Currently, effective axonal regeneration along a very long gap in the there is only one commercial cadaveric decellularized nerve rat, paving the way to its translation to the clinic. We have allograft, Axogen ­Avance® (AxoGen Inc., Alachua, FL, USA) also shown that a decellularized xenograft is not adequate to approved for clinical application by the FDA (Kornfeld et al. ensure nerve regeneration due to interspecies differences of 2019, 2021) but the limited evidence of its pro-regenerative the ECM components to guide adequate migration of regen- potential has prevented its expansion to the European mar- erating cells and axons. ket. Therefore, further studies aimed to optimize protocols to 13 362 Cell and Tissue Research (2022) 390:355–366 Fig. 5  Immunohistochemical characterization of the grafts 1 and blasts, and DAPI (c′). Proximal segment of DC-HX at 14 days post 2 weeks after implantation. Representative micrographs show- injury (d‴″) labeling myelinated axons and Schwann cells (d), and ing immunofluorescence of proximal segment of DC-RA at 7 days macrophages, fibroblasts and DAPI (d′). Images (a, a′, b, b′, c, c′, post injury (a‴″) labeling myelinated axons and Schwann cells (a), d, d′) taken at 40 × magnification; scale bar 500 µm. Below images and macrophages, fibroblasts and DAPI (a′). Proximal segment of showed myelinated axons labeled against NF200 (a″, b″, c″, d″), DC-HX at 7 days post injury (b‴″) labeling myelinated axons and Schwann cells labeled against S100 (a‴, b‴, c‴, d‴), macrophages Schwann cells (b), and macrophages, fibroblasts, and DAPI (b′). labeled against Iba1 (a″″, b″″, c″″, d″″), fibroblasts labeled against Proximal segment of DC-RA at 14 days post injury (c‴″) labeling vimentin (a‴″, b‴″, c‴″, d‴″), images taken at 200 × magnification; myelinated axons and Schwann cells (c), and macrophages, fibro- scale bar 200 µm decellularize human cadaveric nerves and to evaluate its regen- while removing all types of cells, as demonstrated by the lack of erative potential compared to the gold standard autografts are DAPI and the Schwann cell marker S100 in both rat and human needed. In this sense, decellularization is a key element when decellularized grafts. using allografts since the presence of cells on these grafts trig- More importantly, this optimized decellularized allograft ger a strong immune reaction in the recipient that limits their allows successful axonal regeneration when used to repair a clinical use. The targets of rejection in peripheral nerves are critical 15 mm nerve defect in the rat sciatic nerve. However, mainly the Schwann cells and the myelin sheaths, because their regeneration and reinnervation, evaluated by both electro- membranes carry the antigens of the major histocompatibility physiological tests and histological techniques, is slower in complex (Evans et al. 1995). Systemic immunosuppression decellularized allografts compared to the native autografts. is effective for avoiding rejection (Udina et al. 2003), but it This finding is not surprising, since the decellularization is not a convenient alternative for this type of non-vital graft, eliminates all cellular components, among them Schwann since it would place the patient at risk of infection, toxicity, cells, that are key elements orchestrating the regenerative and other complications (Szynkaruk et al. 2013; Vasudevan response in the distal stump. Migration of host Schwann et al. 2014). Therefore, the decellularization protocol has to cells into the allograft is needed to create a permissive milieu guarantee elimination of donor antigens while preserving the for regeneration, a fact that explains the slower regeneration ECM and the physical architecture of the nerve, two elements observed into a decellularized graft (Hall 1996). that are fundamental to sustain successful regeneration. In this On the other hand, regeneration and reinnervation were sense, our protocol of decellularization (Nieto-Nicolau et al. more limited when the same gap was repaired using a decel- 2021) successfully preserves the ECM and endoneurial tubules, lularized xenograft, obtained from a human nerve. As previ- visualized by immunolabeling against laminin and collagen IV, ously described (Wood et al. 2014), there were appreciable 13 Cell and Tissue Research (2022) 390:355–366 363 Fig. 6  In vitro neural growth on laminins from different species. Plots DRG neuronal culture immunolabeled against β3-tubulin (green), showing the the neurite length of all neurons (a), the percentage of nf200 (red) and dapi (blue), on surface coated with mouse laminin neurons with neurites (b), the neurite length of myelinated neurons (e) or with human laminin (e′). Higher number of neurites and longer (nf200 +) (c) and non-myelinated (nf200 −) neurons with neurites neurites can be observed on mouse laminin than on human laminin. (d). *p < 0.5 vs mice laminin; **p < 0.01 vs mice laminin. Mouse Images taken at 100 × magnification; scale bar 200 µm qualitative differences in the arrangement of axons (more dis- nerve defect (Kvist et al. 2011), whereas another found mini- organized) and in the density of myelinated axons (reduced) mal differences by using the same gap to repair the rat facial into the graft and distally in the decellularized xenografts nerve (Huang et al. 2015). The origin of the donor could compared to decellularized allografts and more with respect hardly explain these differences, since Kvist et al. compared to the cellular autografts. A previous study also reported the xenografts from different donors and could not found a sim- higher pro-regenerative capability of allografts compared ple relation between the origin of the graft and the extent to to xenografts, when repairing a smaller, 10 mm rat sciatic which it supports axonal outgrowth (Kvist et al. 2011). 13 364 Cell and Tissue Research (2022) 390:355–366 It remains unclear why and how interspecies differences Evercooren et al. 1982). We found that laminin was present between the host and the donor graft affect nerve regenera- in the basal lamina, inside the preserved endoneurial tubules, tion (Kvist et al. 2011; Wood et al. 2014). By using a decel- similarly in both types of decellularized grafts. However, lularized graft, no immune rejection is expected. Therefore, when we compared the capability of mouse neurons to extend we decided to evaluate and compare short-term regeneration neurites in vitro, in wells coated with laminin either from within decellularized allografts and xenografts. Consistent mouse or from human sources, we observed that laminin with the long-term study, we found that regeneration was from the same specie sustained stronger and longer neurite faster in DC-RA group, whereas in the DC-HX group was growth. This intra-specie preference can contribute to the limited or even failed; furthermore, this failure was accompa- higher potential of allografts to sustain axonal growth com- nied with a marked cellular infiltration, evident at two weeks pared to xenografts (Kvist et al. 2011). after grafting, not earlier, suggesting that a period of time is In conclusion, a nerve decellularization protocol that guar- necessary to trigger an inflammatory/rejection response in antees elimination of cellular contents preserving the struc- these grafts, negatively impacting the regenerative process. ture and the ECM of the graft can become an alternative to Interestingly, in the DC-RA grafts, we observed an early autografts to repair long nerve defects, although regeneration fibroblast proliferation, that was organized in aligned tubes is going to be slower in the decellularized grafts. In contrast, inside the nerve at 14 days post grafting. It is well estab- xenografts showed some limitations, mainly related to inter- lished that fibroblasts are key players in the formation of species immune response, fibrotic reaction, and the lower the initial bridge between the two nerve stumps after nerve capacity of ECM components to sustain axon growth of neu- transection (Parrinello et al. 2010). This bridge will be later rons from another specie. Our results demonstrate the transla- invaded by migrating Schwann cells to create the adequate tionality of the optimized protocol for nerve decellularization milieu for axons to regenerate through the graft. In fact, that fulfills the standardized recommendations (Nieto-Nicolau migration of Schwann cells is directed by fibroblasts and et al. 2021), and also its success as a decellularized allograft endothelial cells (Cattin et al. 2015), whereas Schwann cell to sustain regeneration over long nerve gaps in rats. invasion precede growth of the axons through the bridge (Chen et al. 2005; Mcdonald et al. 2006; Parrinello et al. Acknowledgements The authors thank Jessica Jaramillo, Neus Hernández, and Mónica Espejo for technical support. 2010). In the case of a decellularized graft, both fibroblasts and Schwann cells have to migrate from the proximal and Funding Open Access Funding provided by Universitat Autonoma de distal nerve stumps and re-cellularize the graft. Recellu- Barcelona. This study was partially supported by funds from Banc de larization by Schwann cells seems mandatory to guarantee Sang i Teixits, project SAF2017-84464R, from Ministerio de Ciencia, regeneration, but fibroblasts are also important to facilitate Innovación y Universidades of Spain, and CIBERNED (CB06/05/1105) and TERAV (RD21/0017/0008) funds from the Instituto de Salud Car- attraction of Schwann cells. However, when we analyzed los III of Spain, co-funded by European Union (ERDF/ESF). the decellularized grafts at 7dpi, we observed that axons preceded Schwann cells, indicating that the decellularized Declarations graft already offers a permissive milieu for axonal regenera- tion, thanks to the preserved architecture and the presence Ethical approval All human tissue processing and animal experiments of ECM components. In addition, fibroblasts aligned along protocols were approved with CEIm Hospital Valle Hebron, Barcelona; the endoneurial tubules, would present a pro-regenerative PR (BST) 314/2019 and Ethics Committee of the Universitat Autònoma de Barcelona and Generalitat de Catalunya (reference #10306) and profile that favor axonal regeneration and contrast with the were carried out in accordance with national and European guidelines fibrotic reaction observed in other situations, that is det- (EEC regulations 2004/23/CE and 2006/17/CE, European Commu- rimental for regeneration. In fact, in DC-HX grafts, large nity Council Directive 2010/63/EU and Spanish Law 14/2007 and RD number of fibroblasts were also observed, but mainly located 1716/2011) as detailed in the first subsection of methods. around the nerve, in a reaction compatible with a foreign Informed consent For samples from human nerve cadavers, informed body response. Therefore, the different response of fibro- consent was obtained from the donors’ relatives. blasts can contribute to the different regenerative potential of these two types of grafts. Conflict of interest The authors declare no competing interests. On the other hand, the potential inter-species differences in ECM components can also contribute to the poorer regen- Open Access This article is licensed under a Creative Commons Attri- erative potential of a xenograft compared to an allograft. bution 4.0 International License, which permits use, sharing, adapta- tion, distribution and reproduction in any medium or format, as long Laminin is an important component of the ECM that interacts as you give appropriate credit to the original author(s) and the source, with axons and Schwann cells through binding with integ- provide a link to the Creative Commons licence, and indicate if changes rins (Gonzalez-Perez et al. 2013; Mckerracher et al. 1996) were made. The images or other third party material in this article are and that also promotes neurite growth in culture (Baron-Van included in the article's Creative Commons licence, unless indicated 13 Cell and Tissue Research (2022) 390:355–366 365 otherwise in a credit line to the material. If material is not included in Huang H, Xiao H, Liu H, Niu Y, Yan R, Hu M (2015) A comparative the article's Creative Commons licence and your intended use is not study of acellular nerve xenografts and allografts in repairing rat permitted by statutory regulation or exceeds the permitted use, you will facial nerve defects. Mol Med Rep 12(4):6330–6336. https://​doi.​ need to obtain permission directly from the copyright holder. To view a org/​10.​3892/​mmr.​2015.​4123 copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. Hundepool CA, Nijhuis TH, Kotsougiani D, Friedrich PF, Bishop AT, Shin AY (2017) Optimizing decellularization techniques to create a new nerve allograft: an in vitro study using rodent nerve segments. Neurosurg Focus 42(3):E4. https://​doi.​org/​10.​ References 3171/​2017.1.​FOCUS​16462 Kasper M, Deister C, Beck F, Schmidt CE (2020) Bench-to-bedside Alé A, Bruna J, Morell M, van de Velde H, Monbaliu J, Navarro X, lessons learned: commercialization of an acellular nerve graft. Udina E (2014) Treatment with anti-TNF alpha protects against Adv Health Mater 9:e2000174. https://​doi.​org/​10.​1002/​adhm.​ the neuropathy induced by the proteasome inhibitor bortezomib 20200​0174 in a mouse model. Exp Neurol 253:165–173. https://​doi.​org/​10.​ Kim HA, Mindos T, Parkinson DB (2013) Plastic fantastic: Schwann 1016/j.​expne​urol.​2013.​12.​020 cells and repair of the peripheral nervous system. Stem Cells Allodi I, Udina E, Navarro X (2012) Specificity of peripheral nerve Transl Med 2:553–557. https://​doi.​org/​10.​5966/​sctm.​2013-​0011 regeneration: interactions at the axon level. Progr Neurobiol Kornfeld T, Borger A, Radtke C (2021) Reconstruction of critical 98:16–37. https://​doi.​org/​10.​1016/j.​pneur​obio.​2012.​05.​005 nerve defects using allogenic nerve tissue: a review of current Asplund M, Nilsson M, Jacobsson A, von Holst H (2009) Incidence approaches. Int J Mol Sci 22:3515. https://​d oi.​o rg/​1 0.​3 390/​ of traumatic peripheral nerve injuries and amputations in Sweden ijms2​20735​15 between 1998 and 2006. Neuroepidemiology 32:217–228. https://​ Kornfeld T, Vogt PM, Radtke C (2019) Nerve grafting for peripheral doi.​org/​10.​1159/​00019​7900 nerve injuries with extended defect sizes. Wien Med Wochen- Baron-Van Evercooren A, Kleinman HK, Ohno S, Marangos P, schr 169:240–251. https://​doi.​org/​10.​1007/​s10354-​018-​0675-6 Schwartz JP, Dubois-Dalcq ME (1982) Nerve growth factor, Kvist M, Sondell M, Kanje M, Dahlin LB (2011) Regeneration in, laminin, and fibronectin promote neurite growth in human fetal and properties of, extracted peripheral nerve allografts and xeno- sensory ganglia cultures. J Neurosci Res 8:179–193. https://​doi.​ grafts. J Plast Surg Hand Surg 45:122–128. https://​doi.​org/​10.​ org/​10.​1002/​jnr.​49008​0208 3109/​20006​56X.​2011.​571847 Cattin AL, Burden JJ, Van Emmenis L, Mackenzie FE, Hoving Lovati AB, D’Arrigo D, Odella S, Tos P, Geuna S, Raimondo S JJ, Garcia Calavia N, Guo Y, McLaughlin M, Rosenberg LH, (2018) Nerve repair using decellularized nerve grafts in rat Quereda V, Jamecna D, Napoli I, Parrinello S, Enver T, Ruhrberg models. A review of the literature. Front Cell Neurosci 12:427. C, Lloyd AC (2015) Macrophage-induced blood vessels guide https://​doi.​org/​10.​3389/​fncel.​2018.​00427 Schwann cell-mediated regeneration of peripheral nerves. Cell McDonald D, Cheng C, Chen Y, Zochodne D (2006) Early events 162:1127–1139. https://​doi.​org/​10.​1016/j.​cell.​2015.​07.​021 of peripheral nerve regeneration. Neuron Glia Biol 2:139–147. Cattin AL, Lloyd AC (2016) The multicellular complexity of peripheral https://​doi.​org/​10.​1017/​S1740​925X0​50003​47 nerve regeneration. Curr Opin Neurobiol 39:38–46. https://​doi.​ McKerracher L, Chamoux M, Arregui CO (1996) Role of laminin and org/​10.​1016/j.​conb.​2016.​04.​005 integrin interactions in growth cone guidance. Mol Neurobiol Chen YY, McDonald D, Cheng C, Magnowski B, Durand J, Zochodne 12:95–116. https://​doi.​org/​10.​1007/​BF027​40648 DW (2005) Axon and Schwann cell partnership during nerve Nagao RJ, Lundy S, Khaing ZZ, Schmidt CE (2011) Functional charac- regrowth. J Neuropathol Exp Neurol 64:613–622. https://​doi.​org/​ terization of optimized acellular peripheral nerve graft in a rat sciatic 10.​1097/​01.​jnen.​00001​71650.​94341.​46 nerve injury model. Neurol Res 33:600–608. https://​doi.​org/​10.​1179/​ Dun XP, Parkinson DB (2015) Visualizing peripheral nerve regenera- 17431​32810Y.​00000​00023 tion by whole mount staining. PLoS ONE 10:e0119168. https://​ Navarro X, Butí M, Verdú E (1994) Autotomy prevention by amytrip- doi.​org/​10.​1371/​journ​al.​pone.​01191​68 tiline after peripheral nerve section in different strains of mice. Evans PJ, Midha R, Mackinnon SE (1995) The peripheral nerve allo- Restor Neurol Neurosci 6:151–157. https://​doi.​org/​10.​3233/​ graft: a comprehensive review of regeneration and neuroimmu- RNN-​1994-​6209 nology. Prog Neurobiol 43:187–233. https://​doi.​org/​10.​1016/​ Navarro X, Vivó M, Valero-Cabré A (2007) Neural plasticity after 0301-​0082(94)​90001-9 peripheral nerve injury and regeneration. Prog Neurobiol Gaudin R, Knipfer C, Henningsen A, Smeets R, Heiland M, Hadlock 82:163–201. https://​doi.​org/​10.​1016/j.​pneur​obio.​2007.​06.​005 T (2016) Approaches to peripheral nerve repair: generations of Nieto-Nicolau N, López-Chicón P, Fariñas O, Bolívar S, Udina E, biomaterial conduits yielding to replacing autologous nerve grafts Navarro X, Casaroli-Marano R, Vilarrodona A (2021) Effec- in craniomaxillofacial surgery. Biomed Res Int 2016:3856262. tive decellularization of human nerve matrix for regenerative https://​doi.​org/​10.​1155/​2016/​38562​62 medicine with a novel protocol. Cell Tissue Res 384:167–177. Gómez N, Cuadras J, Butí M, Navarro X (1996) Histological assessment https://​doi.​org/​10.​1007/​s00441-​020-​03317-3 of sciatic nerve regeneration following resectiosn and grafts or tube Nieto-Nicolau N, López-Chicón P, Torrico C, Bolívar S, Contreras- repair in the mouse. Restor Neurol Neurosci 10:187–196 Carreton E, Udina E, Navarro X, Casaroli-Marano RP, Fariñas Gonzalez-Perez F, Udina E, Navarro X (2013) Extracellular matrix O, Vilarrodona A (2022) “Off-the-Shelf” nerve matrix preserva- components in peripheral nerve regeneration. Int Rev Neurobiol tion. Biopreserv Biobank 20:48–58. https://​doi.​org/​10.​1089/​bio.​ 108:257–275. https://​d oi.​o rg/​1 0.​1 016/​B 978-0-​1 2-​4 10499-​0 .​ 2020.​0158 (Epub 2021 Sep 20 PMID: 34542324) 00010-1 Parrinello S, Napoli I, Ribeiro S, Wingfield Digby P, Fedorova M, Parkin- Grinsell D, Keating CP (2014) Peripheral nerve reconstruction after son DB, Doddrell RD, Nakayama M, Adams RH, Lloyd AC (2010) injury: a review of clinical and experimental therapies. Biomed EphB signaling directs peripheral nerve regeneration through Sox2- Res Int 2014:698256. https://​doi.​org/​10.​1155/​2014/​698256 dependent Schwann cell sorting. Cell 143:145–155. https://​doi.​org/​10.​ Hall SM (1996) Regeneration in cellular and acellular autografts in the 1016/j.​cell.​2010.​08.​039 (PMID: 20869108; PMCID: PMC3826531) peripheral nervous system. Neuropathol Appl Neurobiol 12:27– Pfister BJ, Gordon T, Loverde JR, Kochar AS, Mackinnon SE, Cullen 46. https://​doi.​org/​10.​1111/j.​1365-​2990.​1986.​tb006​79.x DK (2011) Biomedical engineering strategies for peripheral 13 366 Cell and Tissue Research (2022) 390:355–366 nerve repair: surgical applications, state of the art, and future allografts in peripheral nerve gap reconstruction. Tissue Eng Part challenges. Crit Rev Biomed Eng 39:81–124. https://​doi.​org/​10.​ B Rev 19:83–96. https://​doi.​org/​10.​1089/​ten.​TEB.​2012.​0275 1615/​critr​evbio​medeng.​v39.​i2.​20 Udina E, Voda J, Gold BG, Navarro X (2003) Comparative dose Philips C, Campos F, Roosens A, Sánchez-Quevedo MDC, Declercq dependence study of FK506 on transected mouse sciatic nerve H, Carriel V (2018) Qualitative and quantitative evaluation of repaired by allografts or xenografts. J Periph Nerv Syst 8:145–154 a novel detergent-based method for decellularization of periph- Vasudevan S, Huang J, Botterman B, Matloub HS, Keefer E, Cheng eral nerves. Ann Biomed Eng 46:1921–1937. https://​doi.​org/​10.​ J (2014) Detergent-free decellularized nerve grafts for long-gap 1007/​s10439-​018-​2082-y peripheral nerve reconstruction. Plast Reconstr Surg Glob Open Roballo KCS, Bushman J (2019) Evaluation of the host immune 2:e201. https://​doi.​org/​10.​1097/​GOX.​00000​00000​000118 response and functional recovery in peripheral nerve autografts Wood MD, Kemp SW, Liu EH, Szynkaruk M, Gordon T, Borschel GH and allografts. Transpl Immunol 53:61–71. https://​doi.​org/​10.​ (2014) Rat-derived processed nerve allografts support more axon 1016/j.​trim.​2019.​01.​003 regeneration in rat than human-derived processed nerve xeno- Rodríguez FJ, Verdú E, Ceballos D, Navarro X (2000) Nerve guides grafts. J Biomed Mater Res A 102:1085–1091. https://​doi.​org/​ seeded with autologous schwann cells improve nerve regeneration. 10.​1002/​jbm.a.​34773 Exp Neurol 161:571–584. https://d​ oi.o​ rg/1​ 0.1​ 006/e​ xnr.1​ 999.7​ 315 Szynkaruk M, Kemp SW, Wood MD, Gordon T, Borschel GH (2013) Publisher's Note Springer Nature remains neutral with regard to Experimental and clinical evidence for use of decellularized nerve jurisdictional claims in published maps and institutional affiliations. 13